entry_point
stringlengths 1
65
| original_triton_code
stringlengths 4.5k
619k
| python_code
stringlengths 208
60.9k
| triton_code
stringlengths 1.15k
275k
| repo_name
stringlengths 7
115
| module_name
stringlengths 1
65
| synthetic
bool 1
class | uuid
int64 0
18.5k
| licenses
sequencelengths 1
6
| stars
int64 0
19.8k
| sha
stringlengths 40
40
| repo_link
stringlengths 72
180
| pytorch_code
stringlengths 200
4.05k
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|
Conv2dSamePadding | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/jw/cjw2limywkgcxcmakwde6yhejpgbx3f2uhb57pqab2yabyayyfay.py
# Topologically Sorted Source Nodes: [input_1], Original ATen: [aten.constant_pad_nd]
# Source node to ATen node mapping:
# input_1 => constant_pad_nd
# Graph fragment:
# %constant_pad_nd : [num_users=2] = call_function[target=torch.ops.aten.constant_pad_nd.default](args = (%primals_3, [0, 1, 0, 1], 0.0), kwargs = {})
triton_poi_fused_constant_pad_nd_0 = async_compile.triton('triton_poi_fused_constant_pad_nd_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[512],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_constant_pad_nd_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_constant_pad_nd_0(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 400
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = (xindex // 5) % 5
x0 = xindex % 5
x2 = (xindex // 25)
x3 = xindex
tmp0 = x1
tmp1 = tl.full([1], 4, tl.int64)
tmp2 = tmp0 < tmp1
tmp3 = x0
tmp4 = tmp3 < tmp1
tmp5 = tmp2 & tmp4
tmp6 = tl.load(in_ptr0 + (x0 + (4*x1) + (16*x2)), tmp5 & xmask, other=0.0)
tl.store(out_ptr0 + (x3), tmp6, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/32/c32v7egt4mupqssam3gmac2qgv3ujprjybthsgweflmot256qqw7.py
# Topologically Sorted Source Nodes: [conv2d], Original ATen: [aten.convolution]
# Source node to ATen node mapping:
# conv2d => convolution
# Graph fragment:
# %convolution : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%constant_pad_nd, %primals_1, %primals_2, [1, 1], [1, 1], [1, 1], False, [0, 0], 1), kwargs = {})
triton_poi_fused_convolution_1 = async_compile.triton('triton_poi_fused_convolution_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_1', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_1(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = (xindex // 16) % 4
tmp0 = tl.load(in_out_ptr0 + (x3), xmask)
tmp1 = tl.load(in_ptr0 + (x1), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + (x3), tmp2, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (4, ), (1, ))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 5, 5), (100, 25, 5, 1), torch.float32)
# Topologically Sorted Source Nodes: [input_1], Original ATen: [aten.constant_pad_nd]
stream0 = get_raw_stream(0)
triton_poi_fused_constant_pad_nd_0.run(primals_3, buf0, 400, grid=grid(400), stream=stream0)
del primals_3
# Topologically Sorted Source Nodes: [conv2d], Original ATen: [aten.convolution]
buf1 = extern_kernels.convolution(buf0, primals_1, stride=(1, 1), padding=(1, 1), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf1, (4, 4, 4, 4), (64, 16, 4, 1))
buf2 = buf1; del buf1 # reuse
# Topologically Sorted Source Nodes: [conv2d], Original ATen: [aten.convolution]
triton_poi_fused_convolution_1.run(buf2, primals_2, 256, grid=grid(256), stream=stream0)
del primals_2
return (buf2, primals_1, buf0, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
from torch import nn
import torch.nn.functional as F
def conv2d_same_padding(input, weight, bias=None, stride=1, dilation=1,
groups=1):
input_rows = input.size(2)
filter_rows = weight.size(2)
effective_filter_size_rows = (filter_rows - 1) * dilation[0] + 1
out_rows = (input_rows + stride[0] - 1) // stride[0]
padding_rows = max(0, (out_rows - 1) * stride[0] +
effective_filter_size_rows - input_rows)
rows_odd = padding_rows % 2 != 0
input_cols = input.size(3)
filter_cols = weight.size(3)
effective_filter_size_cols = (filter_cols - 1) * dilation[1] + 1
out_cols = (input_cols + stride[1] - 1) // stride[1]
padding_cols = max(0, (out_cols - 1) * stride[1] +
effective_filter_size_cols - input_cols)
cols_odd = padding_cols % 2 != 0
if rows_odd or cols_odd:
input = F.pad(input, [0, int(cols_odd), 0, int(rows_odd)])
return F.conv2d(input, weight, bias, stride, padding=(padding_rows // 2,
padding_cols // 2), dilation=dilation, groups=groups)
class Conv2dSamePadding(nn.Conv2d):
"""Represents the "Same" padding functionality from Tensorflow.
See: https://github.com/pytorch/pytorch/issues/3867
This solution is mostly copied from
https://github.com/pytorch/pytorch/issues/3867#issuecomment-349279036
Note that the padding argument in the initializer doesn't do anything now
"""
def forward(self, input):
return conv2d_same_padding(input, self.weight, self.bias, self.
stride, self.dilation, self.groups)
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'in_channels': 4, 'out_channels': 4, 'kernel_size': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch import nn
import torch.nn.functional as F
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_poi_fused_constant_pad_nd_0(in_ptr0, out_ptr0, xnumel, XBLOCK:
tl.constexpr):
xnumel = 400
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = xindex // 5 % 5
x0 = xindex % 5
x2 = xindex // 25
x3 = xindex
tmp0 = x1
tmp1 = tl.full([1], 4, tl.int64)
tmp2 = tmp0 < tmp1
tmp3 = x0
tmp4 = tmp3 < tmp1
tmp5 = tmp2 & tmp4
tmp6 = tl.load(in_ptr0 + (x0 + 4 * x1 + 16 * x2), tmp5 & xmask, other=0.0)
tl.store(out_ptr0 + x3, tmp6, xmask)
@triton.jit
def triton_poi_fused_convolution_1(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl
.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = xindex // 16 % 4
tmp0 = tl.load(in_out_ptr0 + x3, xmask)
tmp1 = tl.load(in_ptr0 + x1, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + x3, tmp2, xmask)
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (4,), (1,))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 5, 5), (100, 25, 5, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_constant_pad_nd_0[grid(400)](primals_3, buf0, 400,
XBLOCK=256, num_warps=4, num_stages=1)
del primals_3
buf1 = extern_kernels.convolution(buf0, primals_1, stride=(1, 1),
padding=(1, 1), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf1, (4, 4, 4, 4), (64, 16, 4, 1))
buf2 = buf1
del buf1
triton_poi_fused_convolution_1[grid(256)](buf2, primals_2, 256,
XBLOCK=256, num_warps=4, num_stages=1)
del primals_2
return buf2, primals_1, buf0
def conv2d_same_padding(input, weight, bias=None, stride=1, dilation=1,
groups=1):
input_rows = input.size(2)
filter_rows = weight.size(2)
effective_filter_size_rows = (filter_rows - 1) * dilation[0] + 1
out_rows = (input_rows + stride[0] - 1) // stride[0]
padding_rows = max(0, (out_rows - 1) * stride[0] +
effective_filter_size_rows - input_rows)
rows_odd = padding_rows % 2 != 0
input_cols = input.size(3)
filter_cols = weight.size(3)
effective_filter_size_cols = (filter_cols - 1) * dilation[1] + 1
out_cols = (input_cols + stride[1] - 1) // stride[1]
padding_cols = max(0, (out_cols - 1) * stride[1] +
effective_filter_size_cols - input_cols)
cols_odd = padding_cols % 2 != 0
if rows_odd or cols_odd:
input = F.pad(input, [0, int(cols_odd), 0, int(rows_odd)])
return F.conv2d(input, weight, bias, stride, padding=(padding_rows // 2,
padding_cols // 2), dilation=dilation, groups=groups)
class Conv2dSamePaddingNew(nn.Conv2d):
"""Represents the "Same" padding functionality from Tensorflow.
See: https://github.com/pytorch/pytorch/issues/3867
This solution is mostly copied from
https://github.com/pytorch/pytorch/issues/3867#issuecomment-349279036
Note that the padding argument in the initializer doesn't do anything now
"""
def forward(self, input_0):
primals_1 = self.weight
primals_2 = self.bias
primals_3 = input_0
output = call([primals_1, primals_2, primals_3])
return output[0]
| hulaba/pycrop-yield-prediction | Conv2dSamePadding | false | 6,830 | [
"MIT"
] | 1 | b4790dc2f87a73e8a0604e8c22466314090c5abf | https://github.com/hulaba/pycrop-yield-prediction/tree/b4790dc2f87a73e8a0604e8c22466314090c5abf | import torch
from torch import nn
import torch.nn.functional as F
def conv2d_same_padding(input, weight, bias=None, stride=1, dilation=1,
groups=1):
input_rows = input.size(2)
filter_rows = weight.size(2)
effective_filter_size_rows = (filter_rows - 1) * dilation[0] + 1
out_rows = (input_rows + stride[0] - 1) // stride[0]
padding_rows = max(0, (out_rows - 1) * stride[0] +
effective_filter_size_rows - input_rows)
rows_odd = padding_rows % 2 != 0
input_cols = input.size(3)
filter_cols = weight.size(3)
effective_filter_size_cols = (filter_cols - 1) * dilation[1] + 1
out_cols = (input_cols + stride[1] - 1) // stride[1]
padding_cols = max(0, (out_cols - 1) * stride[1] +
effective_filter_size_cols - input_cols)
cols_odd = padding_cols % 2 != 0
if rows_odd or cols_odd:
input = F.pad(input, [0, int(cols_odd), 0, int(rows_odd)])
return F.conv2d(input, weight, bias, stride, padding=(padding_rows // 2,
padding_cols // 2), dilation=dilation, groups=groups)
class Model(nn.Conv2d):
"""Represents the "Same" padding functionality from Tensorflow.
See: https://github.com/pytorch/pytorch/issues/3867
This solution is mostly copied from
https://github.com/pytorch/pytorch/issues/3867#issuecomment-349279036
Note that the padding argument in the initializer doesn't do anything now
"""
def forward(self, input):
return conv2d_same_padding(input, self.weight, self.bias, self.
stride, self.dilation, self.groups)
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4, 4, 4]
|
GCNModelVAE | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/3w/c3w6ed26vbtzgzigkkyuiuwphm777x7aoogcxkvoly3ia7qifnzg.py
# Topologically Sorted Source Nodes: [output_1], Original ATen: [aten.relu]
# Source node to ATen node mapping:
# output_1 => relu
# Graph fragment:
# %relu : [num_users=3] = call_function[target=torch.ops.aten.relu.default](args = (%mm_1,), kwargs = {})
triton_poi_fused_relu_0 = async_compile.triton('triton_poi_fused_relu_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_relu_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_relu_0(in_out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_out_ptr0 + (x0), xmask)
tmp1 = tl.full([1], 0, tl.int32)
tmp2 = triton_helpers.maximum(tmp1, tmp0)
tl.store(in_out_ptr0 + (x0), tmp2, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5 = args
args.clear()
assert_size_stride(primals_1, (4, 4), (4, 1))
assert_size_stride(primals_2, (4, 4), (4, 1))
assert_size_stride(primals_3, (4, 4), (4, 1))
assert_size_stride(primals_4, (4, 4), (4, 1))
assert_size_stride(primals_5, (4, 4), (4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [support], Original ATen: [aten.mm]
extern_kernels.mm(primals_1, primals_2, out=buf0)
del primals_2
buf1 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [output], Original ATen: [aten.mm]
extern_kernels.mm(primals_3, buf0, out=buf1)
buf2 = buf1; del buf1 # reuse
# Topologically Sorted Source Nodes: [output_1], Original ATen: [aten.relu]
stream0 = get_raw_stream(0)
triton_poi_fused_relu_0.run(buf2, 16, grid=grid(16), stream=stream0)
buf3 = buf0; del buf0 # reuse
# Topologically Sorted Source Nodes: [support_1], Original ATen: [aten.mm]
extern_kernels.mm(buf2, primals_4, out=buf3)
buf4 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [output_2], Original ATen: [aten.mm]
extern_kernels.mm(primals_3, buf3, out=buf4)
buf5 = buf3; del buf3 # reuse
# Topologically Sorted Source Nodes: [support_2], Original ATen: [aten.mm]
extern_kernels.mm(buf2, primals_5, out=buf5)
buf6 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [output_3], Original ATen: [aten.mm]
extern_kernels.mm(primals_3, buf5, out=buf6)
del buf5
return (buf4, buf6, buf2, reinterpret_tensor(primals_3, (4, 4), (1, 4), 0), reinterpret_tensor(primals_5, (4, 4), (1, 4), 0), reinterpret_tensor(primals_4, (4, 4), (1, 4), 0), reinterpret_tensor(primals_1, (4, 4), (1, 4), 0), )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| from torch.nn import Module
import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.nn.modules.module import Module
from torch.nn.parameter import Parameter
import torch.nn.modules.loss
class GraphConvolution(Module):
"""
Simple GCN layer, similar to https://arxiv.org/abs/1609.02907
"""
def __init__(self, in_features, out_features, dropout=0.0, act=F.relu):
super(GraphConvolution, self).__init__()
self.in_features = in_features
self.out_features = out_features
self.dropout = dropout
self.act = act
self.weight = Parameter(torch.FloatTensor(in_features, out_features))
self.reset_parameters()
def reset_parameters(self):
torch.nn.init.xavier_uniform_(self.weight)
def forward(self, input, adj):
input = F.dropout(input, self.dropout, self.training)
support = torch.mm(input, self.weight)
output = torch.spmm(adj, support)
output = self.act(output)
return output
def __repr__(self):
return self.__class__.__name__ + ' (' + str(self.in_features
) + ' -> ' + str(self.out_features) + ')'
class InnerProductDecoder(nn.Module):
"""Decoder for using inner product for prediction."""
def __init__(self, dropout, act=torch.sigmoid):
super(InnerProductDecoder, self).__init__()
self.dropout = dropout
self.act = act
def forward(self, z):
z = F.dropout(z, self.dropout, training=self.training)
adj = self.act(torch.mm(z, z.t()))
return adj
class GCNModelVAE(nn.Module):
def __init__(self, input_feat_dim, hidden_dim1, hidden_dim2, dropout):
super(GCNModelVAE, self).__init__()
self.gc1 = GraphConvolution(input_feat_dim, hidden_dim1, dropout,
act=F.relu)
self.gc2 = GraphConvolution(hidden_dim1, hidden_dim2, dropout, act=
lambda x: x)
self.gc3 = GraphConvolution(hidden_dim1, hidden_dim2, dropout, act=
lambda x: x)
self.dc = InnerProductDecoder(dropout, act=lambda x: x)
def encode(self, x, adj):
hidden1 = self.gc1(x, adj)
return self.gc2(hidden1, adj), self.gc3(hidden1, adj)
def reparameterize(self, mu, logvar):
if self.training:
std = torch.exp(logvar)
eps = torch.randn_like(std)
return eps.mul(std).add_(mu)
else:
return mu
def forward(self, x, adj):
mu, logvar = self.encode(x, adj)
z = self.reparameterize(mu, logvar)
return z, mu, logvar
def get_inputs():
return [torch.rand([4, 4]), torch.rand([4, 4])]
def get_init_inputs():
return [[], {'input_feat_dim': 4, 'hidden_dim1': 4, 'hidden_dim2': 4,
'dropout': 0.5}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch.nn import Module
import torch.nn as nn
import torch.nn.functional as F
from torch.nn.modules.module import Module
from torch.nn.parameter import Parameter
import torch.nn.modules.loss
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_relu_0(in_out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_out_ptr0 + x0, xmask)
tmp1 = tl.full([1], 0, tl.int32)
tmp2 = triton_helpers.maximum(tmp1, tmp0)
tl.store(in_out_ptr0 + x0, tmp2, xmask)
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5 = args
args.clear()
assert_size_stride(primals_1, (4, 4), (4, 1))
assert_size_stride(primals_2, (4, 4), (4, 1))
assert_size_stride(primals_3, (4, 4), (4, 1))
assert_size_stride(primals_4, (4, 4), (4, 1))
assert_size_stride(primals_5, (4, 4), (4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.mm(primals_1, primals_2, out=buf0)
del primals_2
buf1 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.mm(primals_3, buf0, out=buf1)
buf2 = buf1
del buf1
get_raw_stream(0)
triton_poi_fused_relu_0[grid(16)](buf2, 16, XBLOCK=16, num_warps=1,
num_stages=1)
buf3 = buf0
del buf0
extern_kernels.mm(buf2, primals_4, out=buf3)
buf4 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.mm(primals_3, buf3, out=buf4)
buf5 = buf3
del buf3
extern_kernels.mm(buf2, primals_5, out=buf5)
buf6 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.mm(primals_3, buf5, out=buf6)
del buf5
return buf4, buf6, buf2, reinterpret_tensor(primals_3, (4, 4), (1, 4), 0
), reinterpret_tensor(primals_5, (4, 4), (1, 4), 0
), reinterpret_tensor(primals_4, (4, 4), (1, 4), 0
), reinterpret_tensor(primals_1, (4, 4), (1, 4), 0)
class GraphConvolution(Module):
"""
Simple GCN layer, similar to https://arxiv.org/abs/1609.02907
"""
def __init__(self, in_features, out_features, dropout=0.0, act=F.relu):
super(GraphConvolution, self).__init__()
self.in_features = in_features
self.out_features = out_features
self.dropout = dropout
self.act = act
self.weight = Parameter(torch.FloatTensor(in_features, out_features))
self.reset_parameters()
def reset_parameters(self):
torch.nn.init.xavier_uniform_(self.weight)
def forward(self, input, adj):
input = F.dropout(input, self.dropout, self.training)
support = torch.mm(input, self.weight)
output = torch.spmm(adj, support)
output = self.act(output)
return output
def __repr__(self):
return self.__class__.__name__ + ' (' + str(self.in_features
) + ' -> ' + str(self.out_features) + ')'
class InnerProductDecoder(nn.Module):
"""Decoder for using inner product for prediction."""
def __init__(self, dropout, act=torch.sigmoid):
super(InnerProductDecoder, self).__init__()
self.dropout = dropout
self.act = act
def forward(self, z):
z = F.dropout(z, self.dropout, training=self.training)
adj = self.act(torch.mm(z, z.t()))
return adj
class GCNModelVAENew(nn.Module):
def __init__(self, input_feat_dim, hidden_dim1, hidden_dim2, dropout):
super(GCNModelVAENew, self).__init__()
self.gc1 = GraphConvolution(input_feat_dim, hidden_dim1, dropout,
act=F.relu)
self.gc2 = GraphConvolution(hidden_dim1, hidden_dim2, dropout, act=
lambda x: x)
self.gc3 = GraphConvolution(hidden_dim1, hidden_dim2, dropout, act=
lambda x: x)
self.dc = InnerProductDecoder(dropout, act=lambda x: x)
def encode(self, x, adj):
hidden1 = self.gc1(x, adj)
return self.gc2(hidden1, adj), self.gc3(hidden1, adj)
def reparameterize(self, mu, logvar):
if self.training:
std = torch.exp(logvar)
eps = torch.randn_like(std)
return eps.mul(std).add_(mu)
else:
return mu
def forward(self, input_0, input_1):
primals_1 = self.gc1.weight
primals_2 = self.gc2.weight
primals_3 = self.gc3.weight
primals_4 = input_0
primals_5 = input_1
output = call([primals_1, primals_2, primals_3, primals_4, primals_5])
return output[0], output[1], output[2]
| hurraygong/scGNN | GCNModelVAE | false | 6,831 | [
"MIT"
] | 1 | bc555895fbd5740ddd82e03187171116889cc10e | https://github.com/hurraygong/scGNN/tree/bc555895fbd5740ddd82e03187171116889cc10e | from torch.nn import Module
import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.nn.modules.module import Module
from torch.nn.parameter import Parameter
import torch.nn.modules.loss
class GraphConvolution(Module):
"""
Simple GCN layer, similar to https://arxiv.org/abs/1609.02907
"""
def __init__(self, in_features, out_features, dropout=0.0, act=F.relu):
super().__init__()
self.in_features = in_features
self.out_features = out_features
self.dropout = dropout
self.act = act
self.weight = Parameter(torch.FloatTensor(in_features, out_features))
self.reset_parameters()
def reset_parameters(self):
torch.nn.init.xavier_uniform_(self.weight)
def forward(self, input, adj):
input = F.dropout(input, self.dropout, self.training)
support = torch.mm(input, self.weight)
output = torch.spmm(adj, support)
output = self.act(output)
return output
def __repr__(self):
return self.__class__.__name__ + ' (' + str(self.in_features
) + ' -> ' + str(self.out_features) + ')'
class InnerProductDecoder(nn.Module):
"""Decoder for using inner product for prediction."""
def __init__(self, dropout, act=torch.sigmoid):
super().__init__()
self.dropout = dropout
self.act = act
def forward(self, z):
z = F.dropout(z, self.dropout, training=self.training)
adj = self.act(torch.mm(z, z.t()))
return adj
class Model(nn.Module):
def __init__(self, input_feat_dim, hidden_dim1, hidden_dim2, dropout):
super().__init__()
self.gc1 = GraphConvolution(input_feat_dim, hidden_dim1, dropout,
act=F.relu)
self.gc2 = GraphConvolution(hidden_dim1, hidden_dim2, dropout, act=
lambda x: x)
self.gc3 = GraphConvolution(hidden_dim1, hidden_dim2, dropout, act=
lambda x: x)
self.dc = InnerProductDecoder(dropout, act=lambda x: x)
def encode(self, x, adj):
hidden1 = self.gc1(x, adj)
return self.gc2(hidden1, adj), self.gc3(hidden1, adj)
def reparameterize(self, mu, logvar):
if self.training:
std = torch.exp(logvar)
eps = torch.randn_like(std)
return eps.mul(std).add_(mu)
else:
return mu
def forward(self, x, adj):
mu, logvar = self.encode(x, adj)
z = self.reparameterize(mu, logvar)
return z, mu, logvar
def get_inputs():
return [torch.rand([4, 4]), torch.rand([4, 4])]
def get_init_inputs():
return [[], {'input_feat_dim': 4, 'hidden_dim1': 4, 'hidden_dim2': 4,
'dropout': 0.5}]
|
_ASPPModule | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/vt/cvtlwsahn5oxfhga3dwjdswt3bhkvbh4zkaufpwgeopelblrwpnw.py
# Topologically Sorted Source Nodes: [pow_1, sum_1, add, pow_2], Original ATen: [aten.pow, aten.sum, aten.add]
# Source node to ATen node mapping:
# add => add
# pow_1 => pow_1
# pow_2 => pow_2
# sum_1 => sum_1
# Graph fragment:
# %pow_1 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%primals_1, 2), kwargs = {})
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%pow_1, [2, 3], True), kwargs = {})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%sum_1, 1e-05), kwargs = {})
# %pow_2 : [num_users=2] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%add, 0.5), kwargs = {})
triton_per_fused_add_pow_sum_0 = async_compile.triton('triton_per_fused_add_pow_sum_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[16, 16],
reduction_hint=ReductionHint.INNER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused_add_pow_sum_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 1, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused_add_pow_sum_0(in_out_ptr0, in_ptr0, xnumel, rnumel, XBLOCK : tl.constexpr):
xnumel = 16
rnumel = 16
RBLOCK: tl.constexpr = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
roffset = 0
rmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
r1 = rindex
x0 = xindex
tmp0 = tl.load(in_ptr0 + (r1 + (16*x0)), xmask, other=0.0)
tmp1 = tmp0 * tmp0
tmp2 = tl.broadcast_to(tmp1, [XBLOCK, RBLOCK])
tmp4 = tl.where(xmask, tmp2, 0)
tmp5 = tl.sum(tmp4, 1)[:, None]
tmp6 = 1e-05
tmp7 = tmp5 + tmp6
tmp8 = libdevice.sqrt(tmp7)
tl.debug_barrier()
tl.store(in_out_ptr0 + (x0), tmp8, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/bk/cbk3oabqcrqeiz2rnrvk4rxkdmk4ef46ye5mdaspijuw6pzegair.py
# Topologically Sorted Source Nodes: [embedding, pow_3, mean, add_1, pow_4], Original ATen: [aten.mul, aten.pow, aten.mean, aten.add]
# Source node to ATen node mapping:
# add_1 => add_1
# embedding => mul
# mean => mean
# pow_3 => pow_3
# pow_4 => pow_4
# Graph fragment:
# %mul : [num_users=2] = call_function[target=torch.ops.aten.mul.Tensor](args = (%pow_2, %primals_2), kwargs = {})
# %pow_3 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%mul, 2), kwargs = {})
# %mean : [num_users=1] = call_function[target=torch.ops.aten.mean.dim](args = (%pow_3, [1], True), kwargs = {})
# %add_1 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mean, 1e-05), kwargs = {})
# %pow_4 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%add_1, 0.5), kwargs = {})
triton_poi_fused_add_mean_mul_pow_1 = async_compile.triton('triton_poi_fused_add_mean_mul_pow_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[4],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_mean_mul_pow_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 8, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_mean_mul_pow_1(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 4
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (4*x0), xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (0))
tmp2 = tl.broadcast_to(tmp1, [XBLOCK])
tmp5 = tl.load(in_ptr0 + (1 + (4*x0)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr1 + (1))
tmp7 = tl.broadcast_to(tmp6, [XBLOCK])
tmp11 = tl.load(in_ptr0 + (2 + (4*x0)), xmask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr1 + (2))
tmp13 = tl.broadcast_to(tmp12, [XBLOCK])
tmp17 = tl.load(in_ptr0 + (3 + (4*x0)), xmask, eviction_policy='evict_last')
tmp18 = tl.load(in_ptr1 + (3))
tmp19 = tl.broadcast_to(tmp18, [XBLOCK])
tmp3 = tmp0 * tmp2
tmp4 = tmp3 * tmp3
tmp8 = tmp5 * tmp7
tmp9 = tmp8 * tmp8
tmp10 = tmp4 + tmp9
tmp14 = tmp11 * tmp13
tmp15 = tmp14 * tmp14
tmp16 = tmp10 + tmp15
tmp20 = tmp17 * tmp19
tmp21 = tmp20 * tmp20
tmp22 = tmp16 + tmp21
tmp23 = 4.0
tmp24 = tmp22 / tmp23
tmp25 = 1e-05
tmp26 = tmp24 + tmp25
tmp27 = libdevice.sqrt(tmp26)
tl.store(out_ptr0 + (x0), tmp27, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/ni/cnixk7pwsdlgsjxwsmpbrzopu3rzfg26mvtp5cemfxo35lnktysi.py
# Topologically Sorted Source Nodes: [embedding, pow_3, mean, add_1, pow_4, norm, mul_1, add_2, tanh, gate], Original ATen: [aten.mul, aten.pow, aten.mean, aten.add, aten.div, aten.tanh]
# Source node to ATen node mapping:
# add_1 => add_1
# add_2 => add_2
# embedding => mul
# gate => add_3
# mean => mean
# mul_1 => mul_1
# norm => div
# pow_3 => pow_3
# pow_4 => pow_4
# tanh => tanh
# Graph fragment:
# %mul : [num_users=2] = call_function[target=torch.ops.aten.mul.Tensor](args = (%pow_2, %primals_2), kwargs = {})
# %pow_3 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%mul, 2), kwargs = {})
# %mean : [num_users=1] = call_function[target=torch.ops.aten.mean.dim](args = (%pow_3, [1], True), kwargs = {})
# %add_1 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mean, 1e-05), kwargs = {})
# %pow_4 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%add_1, 0.5), kwargs = {})
# %div : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%primals_3, %pow_4), kwargs = {})
# %mul_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%mul, %div), kwargs = {})
# %add_2 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul_1, %primals_4), kwargs = {})
# %tanh : [num_users=1] = call_function[target=torch.ops.aten.tanh.default](args = (%add_2,), kwargs = {})
# %add_3 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%tanh, 1.0), kwargs = {})
triton_poi_fused_add_div_mean_mul_pow_tanh_2 = async_compile.triton('triton_poi_fused_add_div_mean_mul_pow_tanh_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: '*fp32', 5: '*fp32', 6: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5, 6), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_div_mean_mul_pow_tanh_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_div_mean_mul_pow_tanh_2(in_ptr0, in_ptr1, in_ptr2, in_ptr3, in_ptr4, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr1 + (x0), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr2 + (x0), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr3 + (x1), xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr4 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 * tmp1
tmp5 = tmp3 / tmp4
tmp6 = tmp2 * tmp5
tmp8 = tmp6 + tmp7
tmp9 = libdevice.tanh(tmp8)
tmp10 = 1.0
tmp11 = tmp9 + tmp10
tl.store(out_ptr0 + (x2), tmp11, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/vz/cvzd7uxhjcuuvhocpajasfl7npgg3w4tjtzribzneczbe5yv6xr6.py
# Topologically Sorted Source Nodes: [embedding, pow_3, mean, add_1, pow_4, norm, mul_1, add_2, tanh, gate, x], Original ATen: [aten.mul, aten.pow, aten.mean, aten.add, aten.div, aten.tanh]
# Source node to ATen node mapping:
# add_1 => add_1
# add_2 => add_2
# embedding => mul
# gate => add_3
# mean => mean
# mul_1 => mul_1
# norm => div
# pow_3 => pow_3
# pow_4 => pow_4
# tanh => tanh
# x => mul_2
# Graph fragment:
# %mul : [num_users=2] = call_function[target=torch.ops.aten.mul.Tensor](args = (%pow_2, %primals_2), kwargs = {})
# %pow_3 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%mul, 2), kwargs = {})
# %mean : [num_users=1] = call_function[target=torch.ops.aten.mean.dim](args = (%pow_3, [1], True), kwargs = {})
# %add_1 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mean, 1e-05), kwargs = {})
# %pow_4 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%add_1, 0.5), kwargs = {})
# %div : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%primals_3, %pow_4), kwargs = {})
# %mul_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%mul, %div), kwargs = {})
# %add_2 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul_1, %primals_4), kwargs = {})
# %tanh : [num_users=1] = call_function[target=torch.ops.aten.tanh.default](args = (%add_2,), kwargs = {})
# %add_3 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%tanh, 1.0), kwargs = {})
# %mul_2 : [num_users=2] = call_function[target=torch.ops.aten.mul.Tensor](args = (%primals_1, %add_3), kwargs = {})
triton_poi_fused_add_div_mean_mul_pow_tanh_3 = async_compile.triton('triton_poi_fused_add_div_mean_mul_pow_tanh_3', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_div_mean_mul_pow_tanh_3', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_div_mean_mul_pow_tanh_3(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 16)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr1 + (x1), xmask, eviction_policy='evict_last')
tmp2 = tmp0 * tmp1
tl.store(out_ptr0 + (x2), tmp2, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/u4/cu4g5g7yw4hmus54eigq2n2o2g7qmofps57iinixiiyqxgzkpd6y.py
# Topologically Sorted Source Nodes: [x_2, relu], Original ATen: [aten.native_group_norm, aten.relu, aten.threshold_backward]
# Source node to ATen node mapping:
# relu => relu
# x_2 => add_4, add_5, mul_4, rsqrt, var_mean
# Graph fragment:
# %var_mean : [num_users=2] = call_function[target=torch.ops.aten.var_mean.correction](args = (%view, [2, 3]), kwargs = {correction: 0, keepdim: True})
# %add_4 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%getitem, 1e-05), kwargs = {})
# %rsqrt : [num_users=2] = call_function[target=torch.ops.aten.rsqrt.default](args = (%add_4,), kwargs = {})
# %mul_4 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%view_1, %unsqueeze_5), kwargs = {})
# %add_5 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul_4, %unsqueeze_2), kwargs = {})
# %relu : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%add_5,), kwargs = {})
# %le : [num_users=1] = call_function[target=torch.ops.aten.le.Scalar](args = (%relu, 0), kwargs = {})
triton_per_fused_native_group_norm_relu_threshold_backward_4 = async_compile.triton('triton_per_fused_native_group_norm_relu_threshold_backward_4', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[4, 512],
reduction_hint=ReductionHint.INNER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: '*fp32', 5: '*i1', 6: '*fp32', 7: 'i32', 8: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5, 6), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused_native_group_norm_relu_threshold_backward_4', 'mutated_arg_names': [], 'no_x_dim': True, 'num_load': 3, 'num_reduction': 4, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused_native_group_norm_relu_threshold_backward_4(in_ptr0, in_ptr1, in_ptr2, out_ptr0, out_ptr2, out_ptr3, out_ptr4, xnumel, rnumel):
xnumel = 4
XBLOCK: tl.constexpr = 1
rnumel = 324
RBLOCK: tl.constexpr = 512
xoffset = tl.program_id(0) * XBLOCK
xindex = tl.full([1], xoffset, tl.int32)
xmask = tl.full([RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[:]
roffset = 0
rmask = rindex < rnumel
r1 = rindex
x0 = xindex
r3 = (rindex // 81)
tmp0 = tl.load(in_ptr0 + (r1 + (324*x0)), rmask, other=0.0)
tmp24 = tl.load(in_ptr1 + (r3), rmask, eviction_policy='evict_last', other=0.0)
tmp26 = tl.load(in_ptr2 + (r3), rmask, eviction_policy='evict_last', other=0.0)
tmp1 = tl.broadcast_to(tmp0, [RBLOCK])
tmp3 = tl.where(rmask, tmp1, 0)
tmp4 = tl.broadcast_to(tmp1, [RBLOCK])
tmp6 = tl.where(rmask, tmp4, 0)
tmp7 = triton_helpers.promote_to_tensor(tl.sum(tmp6, 0))
tmp8 = tl.full([1], 324, tl.int32)
tmp9 = tmp8.to(tl.float32)
tmp10 = tmp7 / tmp9
tmp11 = tmp1 - tmp10
tmp12 = tmp11 * tmp11
tmp13 = tl.broadcast_to(tmp12, [RBLOCK])
tmp15 = tl.where(rmask, tmp13, 0)
tmp16 = triton_helpers.promote_to_tensor(tl.sum(tmp15, 0))
tmp17 = tmp0 - tmp10
tmp18 = 324.0
tmp19 = tmp16 / tmp18
tmp20 = 1e-05
tmp21 = tmp19 + tmp20
tmp22 = libdevice.rsqrt(tmp21)
tmp23 = tmp17 * tmp22
tmp25 = tmp23 * tmp24
tmp27 = tmp25 + tmp26
tmp28 = tl.full([1], 0, tl.int32)
tmp29 = triton_helpers.maximum(tmp28, tmp27)
tmp30 = 0.0
tmp31 = tmp29 <= tmp30
tl.store(out_ptr2 + (r1 + (324*x0)), tmp29, rmask)
tl.store(out_ptr3 + (r1 + (324*x0)), tmp31, rmask)
tl.store(out_ptr4 + (x0), tmp22, None)
tl.store(out_ptr0 + (x0), tmp10, None)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (1, 4, 1, 1), (4, 1, 1, 1))
assert_size_stride(primals_3, (1, 4, 1, 1), (4, 1, 1, 1))
assert_size_stride(primals_4, (1, 4, 1, 1), (4, 1, 1, 1))
assert_size_stride(primals_5, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_6, (4, ), (1, ))
assert_size_stride(primals_7, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 1, 1), (4, 1, 16, 16), torch.float32)
buf1 = reinterpret_tensor(buf0, (4, 4, 1, 1), (4, 1, 1, 1), 0); del buf0 # reuse
# Topologically Sorted Source Nodes: [pow_1, sum_1, add, pow_2], Original ATen: [aten.pow, aten.sum, aten.add]
stream0 = get_raw_stream(0)
triton_per_fused_add_pow_sum_0.run(buf1, primals_1, 16, 16, grid=grid(16), stream=stream0)
buf2 = empty_strided_cuda((4, 1, 1, 1), (1, 4, 4, 4), torch.float32)
# Topologically Sorted Source Nodes: [embedding, pow_3, mean, add_1, pow_4], Original ATen: [aten.mul, aten.pow, aten.mean, aten.add]
triton_poi_fused_add_mean_mul_pow_1.run(buf1, primals_2, buf2, 4, grid=grid(4), stream=stream0)
buf3 = empty_strided_cuda((4, 4, 1, 1), (4, 1, 16, 16), torch.float32)
# Topologically Sorted Source Nodes: [embedding, pow_3, mean, add_1, pow_4, norm, mul_1, add_2, tanh, gate], Original ATen: [aten.mul, aten.pow, aten.mean, aten.add, aten.div, aten.tanh]
triton_poi_fused_add_div_mean_mul_pow_tanh_2.run(buf1, primals_2, primals_3, buf2, primals_4, buf3, 16, grid=grid(16), stream=stream0)
buf4 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [embedding, pow_3, mean, add_1, pow_4, norm, mul_1, add_2, tanh, gate, x], Original ATen: [aten.mul, aten.pow, aten.mean, aten.add, aten.div, aten.tanh]
triton_poi_fused_add_div_mean_mul_pow_tanh_3.run(primals_1, buf3, buf4, 256, grid=grid(256), stream=stream0)
del buf3
# Topologically Sorted Source Nodes: [x_1], Original ATen: [aten.convolution]
buf5 = extern_kernels.convolution(buf4, primals_5, stride=(1, 1), padding=(4, 4), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf5, (4, 4, 9, 9), (324, 81, 9, 1))
buf6 = buf2; del buf2 # reuse
buf10 = empty_strided_cuda((4, 4, 9, 9), (324, 81, 9, 1), torch.float32)
buf11 = empty_strided_cuda((4, 4, 9, 9), (324, 81, 9, 1), torch.bool)
buf9 = empty_strided_cuda((4, 1, 1, 1), (1, 4, 4, 4), torch.float32)
# Topologically Sorted Source Nodes: [x_2, relu], Original ATen: [aten.native_group_norm, aten.relu, aten.threshold_backward]
triton_per_fused_native_group_norm_relu_threshold_backward_4.run(buf5, primals_6, primals_7, buf6, buf10, buf11, buf9, 4, 324, grid=grid(4), stream=stream0)
del primals_7
return (buf10, primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, buf1, buf4, buf5, reinterpret_tensor(buf6, (4, 1), (1, 1), 0), reinterpret_tensor(buf9, (4, 1), (1, 1), 0), buf11, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((1, 4, 1, 1), (4, 1, 1, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((1, 4, 1, 1), (4, 1, 1, 1), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((1, 4, 1, 1), (4, 1, 1, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
class GCT(nn.Module):
def __init__(self, num_channels, epsilon=1e-05, mode='l2', after_relu=False
):
super(GCT, self).__init__()
self.alpha = nn.Parameter(torch.ones(1, num_channels, 1, 1))
self.gamma = nn.Parameter(torch.zeros(1, num_channels, 1, 1))
self.beta = nn.Parameter(torch.zeros(1, num_channels, 1, 1))
self.epsilon = epsilon
self.mode = mode
self.after_relu = after_relu
def forward(self, x):
if self.mode == 'l2':
embedding = (x.pow(2).sum((2, 3), keepdim=True) + self.epsilon
).pow(0.5) * self.alpha
norm = self.gamma / (embedding.pow(2).mean(dim=1, keepdim=True) +
self.epsilon).pow(0.5)
elif self.mode == 'l1':
if not self.after_relu:
_x = torch.abs(x)
else:
_x = x
embedding = _x.sum((2, 3), keepdim=True) * self.alpha
norm = self.gamma / (torch.abs(embedding).mean(dim=1, keepdim=
True) + self.epsilon)
else:
None
exit()
gate = 1.0 + torch.tanh(embedding * norm + self.beta)
return x * gate
class _ASPPModule(nn.Module):
def __init__(self, inplanes, planes, kernel_size, padding, dilation):
super(_ASPPModule, self).__init__()
self.GCT = GCT(inplanes)
self.atrous_conv = nn.Conv2d(inplanes, planes, kernel_size=
kernel_size, stride=1, padding=padding, dilation=dilation, bias
=False)
self.bn = nn.GroupNorm(int(planes / 4), planes)
self.relu = nn.ReLU(inplace=True)
self._init_weight()
def forward(self, x):
x = self.GCT(x)
x = self.atrous_conv(x)
x = self.bn(x)
return self.relu(x)
def _init_weight(self):
for m in self.modules():
if isinstance(m, nn.Conv2d):
torch.nn.init.kaiming_normal_(m.weight)
elif isinstance(m, nn.BatchNorm2d):
m.weight.data.fill_(1)
m.bias.data.zero_()
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'inplanes': 4, 'planes': 4, 'kernel_size': 4, 'padding': 4,
'dilation': 1}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import libdevice
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_per_fused_add_pow_sum_0(in_out_ptr0, in_ptr0, xnumel, rnumel,
XBLOCK: tl.constexpr):
xnumel = 16
RBLOCK: tl.constexpr = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
r1 = rindex
x0 = xindex
tmp0 = tl.load(in_ptr0 + (r1 + 16 * x0), xmask, other=0.0)
tmp1 = tmp0 * tmp0
tmp2 = tl.broadcast_to(tmp1, [XBLOCK, RBLOCK])
tmp4 = tl.where(xmask, tmp2, 0)
tmp5 = tl.sum(tmp4, 1)[:, None]
tmp6 = 1e-05
tmp7 = tmp5 + tmp6
tmp8 = libdevice.sqrt(tmp7)
tl.debug_barrier()
tl.store(in_out_ptr0 + x0, tmp8, xmask)
@triton.jit
def triton_poi_fused_add_mean_mul_pow_1(in_ptr0, in_ptr1, out_ptr0, xnumel,
XBLOCK: tl.constexpr):
xnumel = 4
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + 4 * x0, xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + 0)
tmp2 = tl.broadcast_to(tmp1, [XBLOCK])
tmp5 = tl.load(in_ptr0 + (1 + 4 * x0), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr1 + 1)
tmp7 = tl.broadcast_to(tmp6, [XBLOCK])
tmp11 = tl.load(in_ptr0 + (2 + 4 * x0), xmask, eviction_policy='evict_last'
)
tmp12 = tl.load(in_ptr1 + 2)
tmp13 = tl.broadcast_to(tmp12, [XBLOCK])
tmp17 = tl.load(in_ptr0 + (3 + 4 * x0), xmask, eviction_policy='evict_last'
)
tmp18 = tl.load(in_ptr1 + 3)
tmp19 = tl.broadcast_to(tmp18, [XBLOCK])
tmp3 = tmp0 * tmp2
tmp4 = tmp3 * tmp3
tmp8 = tmp5 * tmp7
tmp9 = tmp8 * tmp8
tmp10 = tmp4 + tmp9
tmp14 = tmp11 * tmp13
tmp15 = tmp14 * tmp14
tmp16 = tmp10 + tmp15
tmp20 = tmp17 * tmp19
tmp21 = tmp20 * tmp20
tmp22 = tmp16 + tmp21
tmp23 = 4.0
tmp24 = tmp22 / tmp23
tmp25 = 1e-05
tmp26 = tmp24 + tmp25
tmp27 = libdevice.sqrt(tmp26)
tl.store(out_ptr0 + x0, tmp27, xmask)
@triton.jit
def triton_poi_fused_add_div_mean_mul_pow_tanh_2(in_ptr0, in_ptr1, in_ptr2,
in_ptr3, in_ptr4, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr1 + x0, xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr2 + x0, xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr3 + x1, xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr4 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 * tmp1
tmp5 = tmp3 / tmp4
tmp6 = tmp2 * tmp5
tmp8 = tmp6 + tmp7
tmp9 = libdevice.tanh(tmp8)
tmp10 = 1.0
tmp11 = tmp9 + tmp10
tl.store(out_ptr0 + x2, tmp11, xmask)
@triton.jit
def triton_poi_fused_add_div_mean_mul_pow_tanh_3(in_ptr0, in_ptr1, out_ptr0,
xnumel, XBLOCK: tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 16
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr1 + x1, xmask, eviction_policy='evict_last')
tmp2 = tmp0 * tmp1
tl.store(out_ptr0 + x2, tmp2, xmask)
@triton.jit
def triton_per_fused_native_group_norm_relu_threshold_backward_4(in_ptr0,
in_ptr1, in_ptr2, out_ptr0, out_ptr2, out_ptr3, out_ptr4, xnumel, rnumel):
XBLOCK: tl.constexpr = 1
rnumel = 324
RBLOCK: tl.constexpr = 512
xoffset = tl.program_id(0) * XBLOCK
xindex = tl.full([1], xoffset, tl.int32)
tl.full([RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[:]
rmask = rindex < rnumel
r1 = rindex
x0 = xindex
r3 = rindex // 81
tmp0 = tl.load(in_ptr0 + (r1 + 324 * x0), rmask, other=0.0)
tmp24 = tl.load(in_ptr1 + r3, rmask, eviction_policy='evict_last',
other=0.0)
tmp26 = tl.load(in_ptr2 + r3, rmask, eviction_policy='evict_last',
other=0.0)
tmp1 = tl.broadcast_to(tmp0, [RBLOCK])
tl.where(rmask, tmp1, 0)
tmp4 = tl.broadcast_to(tmp1, [RBLOCK])
tmp6 = tl.where(rmask, tmp4, 0)
tmp7 = triton_helpers.promote_to_tensor(tl.sum(tmp6, 0))
tmp8 = tl.full([1], 324, tl.int32)
tmp9 = tmp8.to(tl.float32)
tmp10 = tmp7 / tmp9
tmp11 = tmp1 - tmp10
tmp12 = tmp11 * tmp11
tmp13 = tl.broadcast_to(tmp12, [RBLOCK])
tmp15 = tl.where(rmask, tmp13, 0)
tmp16 = triton_helpers.promote_to_tensor(tl.sum(tmp15, 0))
tmp17 = tmp0 - tmp10
tmp18 = 324.0
tmp19 = tmp16 / tmp18
tmp20 = 1e-05
tmp21 = tmp19 + tmp20
tmp22 = libdevice.rsqrt(tmp21)
tmp23 = tmp17 * tmp22
tmp25 = tmp23 * tmp24
tmp27 = tmp25 + tmp26
tmp28 = tl.full([1], 0, tl.int32)
tmp29 = triton_helpers.maximum(tmp28, tmp27)
tmp30 = 0.0
tmp31 = tmp29 <= tmp30
tl.store(out_ptr2 + (r1 + 324 * x0), tmp29, rmask)
tl.store(out_ptr3 + (r1 + 324 * x0), tmp31, rmask)
tl.store(out_ptr4 + x0, tmp22, None)
tl.store(out_ptr0 + x0, tmp10, None)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7) = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (1, 4, 1, 1), (4, 1, 1, 1))
assert_size_stride(primals_3, (1, 4, 1, 1), (4, 1, 1, 1))
assert_size_stride(primals_4, (1, 4, 1, 1), (4, 1, 1, 1))
assert_size_stride(primals_5, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_6, (4,), (1,))
assert_size_stride(primals_7, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 1, 1), (4, 1, 16, 16), torch.float32)
buf1 = reinterpret_tensor(buf0, (4, 4, 1, 1), (4, 1, 1, 1), 0)
del buf0
get_raw_stream(0)
triton_per_fused_add_pow_sum_0[grid(16)](buf1, primals_1, 16, 16,
XBLOCK=8, num_warps=2, num_stages=1)
buf2 = empty_strided_cuda((4, 1, 1, 1), (1, 4, 4, 4), torch.float32)
triton_poi_fused_add_mean_mul_pow_1[grid(4)](buf1, primals_2, buf2,
4, XBLOCK=4, num_warps=1, num_stages=1)
buf3 = empty_strided_cuda((4, 4, 1, 1), (4, 1, 16, 16), torch.float32)
triton_poi_fused_add_div_mean_mul_pow_tanh_2[grid(16)](buf1,
primals_2, primals_3, buf2, primals_4, buf3, 16, XBLOCK=16,
num_warps=1, num_stages=1)
buf4 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
triton_poi_fused_add_div_mean_mul_pow_tanh_3[grid(256)](primals_1,
buf3, buf4, 256, XBLOCK=256, num_warps=4, num_stages=1)
del buf3
buf5 = extern_kernels.convolution(buf4, primals_5, stride=(1, 1),
padding=(4, 4), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf5, (4, 4, 9, 9), (324, 81, 9, 1))
buf6 = buf2
del buf2
buf10 = empty_strided_cuda((4, 4, 9, 9), (324, 81, 9, 1), torch.float32
)
buf11 = empty_strided_cuda((4, 4, 9, 9), (324, 81, 9, 1), torch.bool)
buf9 = empty_strided_cuda((4, 1, 1, 1), (1, 4, 4, 4), torch.float32)
triton_per_fused_native_group_norm_relu_threshold_backward_4[grid(4)](
buf5, primals_6, primals_7, buf6, buf10, buf11, buf9, 4, 324,
num_warps=4, num_stages=1)
del primals_7
return (buf10, primals_1, primals_2, primals_3, primals_4, primals_5,
primals_6, buf1, buf4, buf5, reinterpret_tensor(buf6, (4, 1), (1, 1
), 0), reinterpret_tensor(buf9, (4, 1), (1, 1), 0), buf11)
class GCT(nn.Module):
def __init__(self, num_channels, epsilon=1e-05, mode='l2', after_relu=False
):
super(GCT, self).__init__()
self.alpha = nn.Parameter(torch.ones(1, num_channels, 1, 1))
self.gamma = nn.Parameter(torch.zeros(1, num_channels, 1, 1))
self.beta = nn.Parameter(torch.zeros(1, num_channels, 1, 1))
self.epsilon = epsilon
self.mode = mode
self.after_relu = after_relu
def forward(self, x):
if self.mode == 'l2':
embedding = (x.pow(2).sum((2, 3), keepdim=True) + self.epsilon
).pow(0.5) * self.alpha
norm = self.gamma / (embedding.pow(2).mean(dim=1, keepdim=True) +
self.epsilon).pow(0.5)
elif self.mode == 'l1':
if not self.after_relu:
_x = torch.abs(x)
else:
_x = x
embedding = _x.sum((2, 3), keepdim=True) * self.alpha
norm = self.gamma / (torch.abs(embedding).mean(dim=1, keepdim=
True) + self.epsilon)
else:
None
exit()
gate = 1.0 + torch.tanh(embedding * norm + self.beta)
return x * gate
class _ASPPModuleNew(nn.Module):
def __init__(self, inplanes, planes, kernel_size, padding, dilation):
super(_ASPPModuleNew, self).__init__()
self.GCT = GCT(inplanes)
self.atrous_conv = nn.Conv2d(inplanes, planes, kernel_size=
kernel_size, stride=1, padding=padding, dilation=dilation, bias
=False)
self.bn = nn.GroupNorm(int(planes / 4), planes)
self.relu = nn.ReLU(inplace=True)
self._init_weight()
def _init_weight(self):
for m in self.modules():
if isinstance(m, nn.Conv2d):
torch.nn.init.kaiming_normal_(m.weight)
elif isinstance(m, nn.BatchNorm2d):
m.weight.data.fill_(1)
m.bias.data.zero_()
def forward(self, input_0):
primals_2 = self.GCT.alpha
primals_3 = self.GCT.gamma
primals_4 = self.GCT.beta
primals_1 = self.atrous_conv.weight
primals_6 = self.bn.weight
primals_7 = self.bn.bias
primals_5 = input_0
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7])
return output[0]
| huanglf714/COMatchNet | _ASPPModule | false | 6,832 | [
"Apache-2.0"
] | 1 | 79023f5be65d354eb9bdac026d7e0d73110bc4aa | https://github.com/huanglf714/COMatchNet/tree/79023f5be65d354eb9bdac026d7e0d73110bc4aa | import torch
import torch.nn as nn
class GCT(nn.Module):
def __init__(self, num_channels, epsilon=1e-05, mode='l2', after_relu=False
):
super().__init__()
self.alpha = nn.Parameter(torch.ones(1, num_channels, 1, 1))
self.gamma = nn.Parameter(torch.zeros(1, num_channels, 1, 1))
self.beta = nn.Parameter(torch.zeros(1, num_channels, 1, 1))
self.epsilon = epsilon
self.mode = mode
self.after_relu = after_relu
def forward(self, x):
if self.mode == 'l2':
embedding = (x.pow(2).sum((2, 3), keepdim=True) + self.epsilon
).pow(0.5) * self.alpha
norm = self.gamma / (embedding.pow(2).mean(dim=1, keepdim=True) +
self.epsilon).pow(0.5)
elif self.mode == 'l1':
if not self.after_relu:
_x = torch.abs(x)
else:
_x = x
embedding = _x.sum((2, 3), keepdim=True) * self.alpha
norm = self.gamma / (torch.abs(embedding).mean(dim=1, keepdim=
True) + self.epsilon)
else:
None
exit()
gate = 1.0 + torch.tanh(embedding * norm + self.beta)
return x * gate
class Model(nn.Module):
def __init__(self, inplanes, planes, kernel_size, padding, dilation):
super().__init__()
self.GCT = GCT(inplanes)
self.atrous_conv = nn.Conv2d(inplanes, planes, kernel_size=
kernel_size, stride=1, padding=padding, dilation=dilation, bias
=False)
self.bn = nn.GroupNorm(int(planes / 4), planes)
self.relu = nn.ReLU(inplace=True)
self._init_weight()
def forward(self, x):
x = self.GCT(x)
x = self.atrous_conv(x)
x = self.bn(x)
return self.relu(x)
def _init_weight(self):
for m in self.modules():
if isinstance(m, nn.Conv2d):
torch.nn.init.kaiming_normal_(m.weight)
elif isinstance(m, nn.BatchNorm2d):
m.weight.data.fill_(1)
m.bias.data.zero_()
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'inplanes': 4, 'planes': 4, 'kernel_size': 4, 'padding': 4,
'dilation': 1}]
|
MyMaxPool1dPadSame | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/hp/chpkuk2yt22ulsalxizwhcnmkoxazbhjr2utm4t2bnt43jtjnopn.py
# Topologically Sorted Source Nodes: [net_1], Original ATen: [aten.max_pool2d_with_indices]
# Source node to ATen node mapping:
# net_1 => _low_memory_max_pool2d_with_offsets
# Graph fragment:
# %_low_memory_max_pool2d_with_offsets : [num_users=1] = call_function[target=torch.ops.prims._low_memory_max_pool2d_with_offsets.default](args = (%unsqueeze, [1, 4], [1, 4], [0, 0], [1, 1], False), kwargs = {})
triton_poi_fused_max_pool2d_with_indices_0 = async_compile.triton('triton_poi_fused_max_pool2d_with_indices_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[4],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_max_pool2d_with_indices_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 4, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_max_pool2d_with_indices_0(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 4
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.full([1], -1, tl.int64)
tmp1 = tl.full([1], 0, tl.int64)
tmp2 = tmp0 >= tmp1
tmp3 = tl.full([1], 4, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = tmp2 & tmp4
tmp6 = tl.load(in_ptr0 + ((-1) + (4*x0)), tmp5 & xmask, eviction_policy='evict_last', other=0.0)
tmp7 = tmp1 >= tmp1
tmp8 = tmp1 < tmp3
tmp9 = tmp7 & tmp8
tmp10 = tl.load(in_ptr0 + (4*x0), tmp9 & xmask, eviction_policy='evict_last', other=0.0)
tmp11 = triton_helpers.maximum(tmp10, tmp6)
tmp12 = tl.full([1], 1, tl.int64)
tmp13 = tmp12 >= tmp1
tmp14 = tmp12 < tmp3
tmp15 = tmp13 & tmp14
tmp16 = tl.load(in_ptr0 + (1 + (4*x0)), tmp15 & xmask, eviction_policy='evict_last', other=0.0)
tmp17 = triton_helpers.maximum(tmp16, tmp11)
tmp18 = tl.full([1], 2, tl.int64)
tmp19 = tmp18 >= tmp1
tmp20 = tmp18 < tmp3
tmp21 = tmp19 & tmp20
tmp22 = tl.load(in_ptr0 + (2 + (4*x0)), tmp21 & xmask, eviction_policy='evict_last', other=0.0)
tmp23 = triton_helpers.maximum(tmp22, tmp17)
tl.store(out_ptr0 + (x0), tmp23, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, = args
args.clear()
assert_size_stride(arg0_1, (4, 4), (4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 1, 1), (1, 1, 1), torch.float32)
# Topologically Sorted Source Nodes: [net_1], Original ATen: [aten.max_pool2d_with_indices]
stream0 = get_raw_stream(0)
triton_poi_fused_max_pool2d_with_indices_0.run(arg0_1, buf0, 4, grid=grid(4), stream=stream0)
del arg0_1
return (reinterpret_tensor(buf0, (4, 1), (1, 1), 0), )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([arg0_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.nn.functional as F
class MyMaxPool1dPadSame(nn.Module):
"""
extend nn.MaxPool1d to support SAME padding
"""
def __init__(self, kernel_size):
super(MyMaxPool1dPadSame, self).__init__()
self.kernel_size = kernel_size
self.stride = 1
self.max_pool = torch.nn.MaxPool1d(kernel_size=self.kernel_size)
def forward(self, x):
net = x
in_dim = net.shape[-1]
out_dim = (in_dim + self.stride - 1) // self.stride
p = max(0, (out_dim - 1) * self.stride + self.kernel_size - in_dim)
pad_left = p // 2
pad_right = p - pad_left
net = F.pad(net, (pad_left, pad_right), 'constant', 0)
net = self.max_pool(net)
return net
def get_inputs():
return [torch.rand([4, 4])]
def get_init_inputs():
return [[], {'kernel_size': 4}]
| import torch
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_max_pool2d_with_indices_0(in_ptr0, out_ptr0, xnumel,
XBLOCK: tl.constexpr):
xnumel = 4
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.full([1], -1, tl.int64)
tmp1 = tl.full([1], 0, tl.int64)
tmp2 = tmp0 >= tmp1
tmp3 = tl.full([1], 4, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = tmp2 & tmp4
tmp6 = tl.load(in_ptr0 + (-1 + 4 * x0), tmp5 & xmask, eviction_policy=
'evict_last', other=0.0)
tmp7 = tmp1 >= tmp1
tmp8 = tmp1 < tmp3
tmp9 = tmp7 & tmp8
tmp10 = tl.load(in_ptr0 + 4 * x0, tmp9 & xmask, eviction_policy=
'evict_last', other=0.0)
tmp11 = triton_helpers.maximum(tmp10, tmp6)
tmp12 = tl.full([1], 1, tl.int64)
tmp13 = tmp12 >= tmp1
tmp14 = tmp12 < tmp3
tmp15 = tmp13 & tmp14
tmp16 = tl.load(in_ptr0 + (1 + 4 * x0), tmp15 & xmask, eviction_policy=
'evict_last', other=0.0)
tmp17 = triton_helpers.maximum(tmp16, tmp11)
tmp18 = tl.full([1], 2, tl.int64)
tmp19 = tmp18 >= tmp1
tmp20 = tmp18 < tmp3
tmp21 = tmp19 & tmp20
tmp22 = tl.load(in_ptr0 + (2 + 4 * x0), tmp21 & xmask, eviction_policy=
'evict_last', other=0.0)
tmp23 = triton_helpers.maximum(tmp22, tmp17)
tl.store(out_ptr0 + x0, tmp23, xmask)
def call(args):
arg0_1, = args
args.clear()
assert_size_stride(arg0_1, (4, 4), (4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 1, 1), (1, 1, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_max_pool2d_with_indices_0[grid(4)](arg0_1, buf0, 4,
XBLOCK=4, num_warps=1, num_stages=1)
del arg0_1
return reinterpret_tensor(buf0, (4, 1), (1, 1), 0),
class MyMaxPool1dPadSameNew(nn.Module):
"""
extend nn.MaxPool1d to support SAME padding
"""
def __init__(self, kernel_size):
super(MyMaxPool1dPadSameNew, self).__init__()
self.kernel_size = kernel_size
self.stride = 1
self.max_pool = torch.nn.MaxPool1d(kernel_size=self.kernel_size)
def forward(self, input_0):
arg0_1 = input_0
output = call([arg0_1])
return output[0]
| huynhnhathao/hum_to_find | MyMaxPool1dPadSame | false | 6,833 | [
"MIT"
] | 1 | a0d7ec4bab1a7e2f7175956ff2721e23e2448840 | https://github.com/huynhnhathao/hum_to_find/tree/a0d7ec4bab1a7e2f7175956ff2721e23e2448840 | import torch
import torch.nn as nn
import torch.nn.functional as F
class Model(nn.Module):
"""
extend nn.MaxPool1d to support SAME padding
"""
def __init__(self, kernel_size):
super().__init__()
self.kernel_size = kernel_size
self.stride = 1
self.max_pool = torch.nn.MaxPool1d(kernel_size=self.kernel_size)
def forward(self, x):
net = x
in_dim = net.shape[-1]
out_dim = (in_dim + self.stride - 1) // self.stride
p = max(0, (out_dim - 1) * self.stride + self.kernel_size - in_dim)
pad_left = p // 2
pad_right = p - pad_left
net = F.pad(net, (pad_left, pad_right), 'constant', 0)
net = self.max_pool(net)
return net
def get_inputs():
return [torch.rand([4, 4])]
def get_init_inputs():
return [4]
|
AE | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/7d/c7dsqsqayeybuug3bhkycussd5no66crxmw5vhg5sdol5tpr6u53.py
# Topologically Sorted Source Nodes: [h1], Original ATen: [aten.relu]
# Source node to ATen node mapping:
# h1 => relu
# Graph fragment:
# %add_tensor_3 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mm_default_3, %primals_3), kwargs = {})
# %relu : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%add_tensor_3,), kwargs = {})
triton_poi_fused_relu_0 = async_compile.triton('triton_poi_fused_relu_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[32768],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_relu_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_relu_0(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 32768
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = tl.full([XBLOCK], True, tl.int1)
x2 = xindex
x0 = xindex % 512
tmp0 = tl.load(in_out_ptr0 + (x2), None)
tmp1 = tl.load(in_ptr0 + (x0), None, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tl.store(in_out_ptr0 + (x2), tmp4, None)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/mw/cmwtuxlzmlnevl4fysm4haikyewbpxicjabj2kvkhka5g6sudjww.py
# Topologically Sorted Source Nodes: [z], Original ATen: [aten.relu]
# Source node to ATen node mapping:
# z => relu_1
# Graph fragment:
# %add_tensor_2 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mm_default_2, %primals_5), kwargs = {})
# %relu_1 : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%add_tensor_2,), kwargs = {})
triton_poi_fused_relu_1 = async_compile.triton('triton_poi_fused_relu_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[8192],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_relu_1', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_relu_1(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 8192
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = tl.full([XBLOCK], True, tl.int1)
x2 = xindex
x0 = xindex % 128
tmp0 = tl.load(in_out_ptr0 + (x2), None)
tmp1 = tl.load(in_ptr0 + (x0), None, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tl.store(in_out_ptr0 + (x2), tmp4, None)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/gm/cgmflgdlpeeb52xctoa47uvw47ycyf7ahlj5wdscxdatpbwcboco.py
# Topologically Sorted Source Nodes: [relu_3], Original ATen: [aten.relu, aten.threshold_backward]
# Source node to ATen node mapping:
# relu_3 => relu_3
# Graph fragment:
# %add_tensor : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mm_default, %primals_9), kwargs = {})
# %relu_3 : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%add_tensor,), kwargs = {})
# %le : [num_users=1] = call_function[target=torch.ops.aten.le.Scalar](args = (%relu_3, 0), kwargs = {})
triton_poi_fused_relu_threshold_backward_2 = async_compile.triton('triton_poi_fused_relu_threshold_backward_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*i1', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_relu_threshold_backward_2', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_relu_threshold_backward_2(in_out_ptr0, in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(in_out_ptr0 + (x2), tmp4, xmask)
tl.store(out_ptr0 + (x2), tmp6, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (512, 4), (4, 1))
assert_size_stride(primals_3, (512, ), (1, ))
assert_size_stride(primals_4, (128, 512), (512, 1))
assert_size_stride(primals_5, (128, ), (1, ))
assert_size_stride(primals_6, (512, 128), (128, 1))
assert_size_stride(primals_7, (512, ), (1, ))
assert_size_stride(primals_8, (4, 512), (512, 1))
assert_size_stride(primals_9, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 512), (512, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(primals_1, (64, 4), (4, 1), 0), reinterpret_tensor(primals_2, (4, 512), (1, 4), 0), out=buf0)
del primals_2
buf1 = buf0; del buf0 # reuse
# Topologically Sorted Source Nodes: [h1], Original ATen: [aten.relu]
stream0 = get_raw_stream(0)
triton_poi_fused_relu_0.run(buf1, primals_3, 32768, grid=grid(32768), stream=stream0)
del primals_3
buf2 = empty_strided_cuda((64, 128), (128, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(buf1, reinterpret_tensor(primals_4, (512, 128), (1, 512), 0), out=buf2)
buf3 = buf2; del buf2 # reuse
# Topologically Sorted Source Nodes: [z], Original ATen: [aten.relu]
triton_poi_fused_relu_1.run(buf3, primals_5, 8192, grid=grid(8192), stream=stream0)
del primals_5
buf4 = empty_strided_cuda((64, 512), (512, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(buf3, reinterpret_tensor(primals_6, (128, 512), (1, 128), 0), out=buf4)
buf5 = buf4; del buf4 # reuse
# Topologically Sorted Source Nodes: [h3], Original ATen: [aten.relu]
triton_poi_fused_relu_0.run(buf5, primals_7, 32768, grid=grid(32768), stream=stream0)
del primals_7
buf6 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(buf5, reinterpret_tensor(primals_8, (512, 4), (1, 512), 0), out=buf6)
buf7 = buf6; del buf6 # reuse
buf8 = empty_strided_cuda((64, 4), (4, 1), torch.bool)
# Topologically Sorted Source Nodes: [relu_3], Original ATen: [aten.relu, aten.threshold_backward]
triton_poi_fused_relu_threshold_backward_2.run(buf7, primals_9, buf8, 256, grid=grid(256), stream=stream0)
del primals_9
return (buf7, buf3, reinterpret_tensor(primals_1, (64, 4), (4, 1), 0), buf1, buf3, buf5, buf8, primals_8, primals_6, primals_4, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((512, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((512, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((128, 512), (512, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((128, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((512, 128), (128, 1), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((512, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_8 = rand_strided((4, 512), (512, 1), device='cuda:0', dtype=torch.float32)
primals_9 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.nn.modules.loss
class AE(nn.Module):
""" Autoencoder for dimensional reduction"""
def __init__(self, dim):
super(AE, self).__init__()
self.dim = dim
self.fc1 = nn.Linear(dim, 512)
self.fc2 = nn.Linear(512, 128)
self.fc3 = nn.Linear(128, 512)
self.fc4 = nn.Linear(512, dim)
def encode(self, x):
h1 = F.relu(self.fc1(x))
return F.relu(self.fc2(h1))
def decode(self, z):
h3 = F.relu(self.fc3(z))
return torch.relu(self.fc4(h3))
def forward(self, x):
z = self.encode(x.view(-1, self.dim))
return self.decode(z), z
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'dim': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
import torch.nn as nn
import torch.nn.functional as F
import torch.nn.modules.loss
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_relu_0(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl.constexpr
):
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
tl.full([XBLOCK], True, tl.int1)
x2 = xindex
x0 = xindex % 512
tmp0 = tl.load(in_out_ptr0 + x2, None)
tmp1 = tl.load(in_ptr0 + x0, None, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tl.store(in_out_ptr0 + x2, tmp4, None)
@triton.jit
def triton_poi_fused_relu_1(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl.constexpr
):
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
tl.full([XBLOCK], True, tl.int1)
x2 = xindex
x0 = xindex % 128
tmp0 = tl.load(in_out_ptr0 + x2, None)
tmp1 = tl.load(in_ptr0 + x0, None, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tl.store(in_out_ptr0 + x2, tmp4, None)
@triton.jit
def triton_poi_fused_relu_threshold_backward_2(in_out_ptr0, in_ptr0,
out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(in_out_ptr0 + x2, tmp4, xmask)
tl.store(out_ptr0 + x2, tmp6, xmask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7, primals_8, primals_9) = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (512, 4), (4, 1))
assert_size_stride(primals_3, (512,), (1,))
assert_size_stride(primals_4, (128, 512), (512, 1))
assert_size_stride(primals_5, (128,), (1,))
assert_size_stride(primals_6, (512, 128), (128, 1))
assert_size_stride(primals_7, (512,), (1,))
assert_size_stride(primals_8, (4, 512), (512, 1))
assert_size_stride(primals_9, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 512), (512, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_1, (64, 4), (4, 1), 0),
reinterpret_tensor(primals_2, (4, 512), (1, 4), 0), out=buf0)
del primals_2
buf1 = buf0
del buf0
get_raw_stream(0)
triton_poi_fused_relu_0[grid(32768)](buf1, primals_3, 32768, XBLOCK
=128, num_warps=4, num_stages=1)
del primals_3
buf2 = empty_strided_cuda((64, 128), (128, 1), torch.float32)
extern_kernels.mm(buf1, reinterpret_tensor(primals_4, (512, 128), (
1, 512), 0), out=buf2)
buf3 = buf2
del buf2
triton_poi_fused_relu_1[grid(8192)](buf3, primals_5, 8192, XBLOCK=
256, num_warps=4, num_stages=1)
del primals_5
buf4 = empty_strided_cuda((64, 512), (512, 1), torch.float32)
extern_kernels.mm(buf3, reinterpret_tensor(primals_6, (128, 512), (
1, 128), 0), out=buf4)
buf5 = buf4
del buf4
triton_poi_fused_relu_0[grid(32768)](buf5, primals_7, 32768, XBLOCK
=128, num_warps=4, num_stages=1)
del primals_7
buf6 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
extern_kernels.mm(buf5, reinterpret_tensor(primals_8, (512, 4), (1,
512), 0), out=buf6)
buf7 = buf6
del buf6
buf8 = empty_strided_cuda((64, 4), (4, 1), torch.bool)
triton_poi_fused_relu_threshold_backward_2[grid(256)](buf7,
primals_9, buf8, 256, XBLOCK=128, num_warps=4, num_stages=1)
del primals_9
return buf7, buf3, reinterpret_tensor(primals_1, (64, 4), (4, 1), 0
), buf1, buf3, buf5, buf8, primals_8, primals_6, primals_4
class AENew(nn.Module):
""" Autoencoder for dimensional reduction"""
def __init__(self, dim):
super(AENew, self).__init__()
self.dim = dim
self.fc1 = nn.Linear(dim, 512)
self.fc2 = nn.Linear(512, 128)
self.fc3 = nn.Linear(128, 512)
self.fc4 = nn.Linear(512, dim)
def encode(self, x):
h1 = F.relu(self.fc1(x))
return F.relu(self.fc2(h1))
def decode(self, z):
h3 = F.relu(self.fc3(z))
return torch.relu(self.fc4(h3))
def forward(self, input_0):
primals_2 = self.fc1.weight
primals_3 = self.fc1.bias
primals_4 = self.fc2.weight
primals_5 = self.fc2.bias
primals_6 = self.fc3.weight
primals_7 = self.fc3.bias
primals_8 = self.fc4.weight
primals_9 = self.fc4.bias
primals_1 = input_0
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7, primals_8, primals_9])
return output[0], output[1]
| hurraygong/scGNN | AE | false | 6,834 | [
"MIT"
] | 1 | bc555895fbd5740ddd82e03187171116889cc10e | https://github.com/hurraygong/scGNN/tree/bc555895fbd5740ddd82e03187171116889cc10e | import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.nn.modules.loss
class Model(nn.Module):
""" Autoencoder for dimensional reduction"""
def __init__(self, dim):
super().__init__()
self.dim = dim
self.fc1 = nn.Linear(dim, 512)
self.fc2 = nn.Linear(512, 128)
self.fc3 = nn.Linear(128, 512)
self.fc4 = nn.Linear(512, dim)
def encode(self, x):
h1 = F.relu(self.fc1(x))
return F.relu(self.fc2(h1))
def decode(self, z):
h3 = F.relu(self.fc3(z))
return torch.relu(self.fc4(h3))
def forward(self, x):
z = self.encode(x.view(-1, self.dim))
return self.decode(z), z
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4]
|
VAE | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/zg/czgortk2i5wcve76tkvnwtnxfqclxq6czamnokck6fe5f5wmcjge.py
# Topologically Sorted Source Nodes: [h1], Original ATen: [aten.relu]
# Source node to ATen node mapping:
# h1 => relu
# Graph fragment:
# %add_tensor_2 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mm_default_2, %primals_3), kwargs = {})
# %relu : [num_users=3] = call_function[target=torch.ops.aten.relu.default](args = (%add_tensor_2,), kwargs = {})
triton_poi_fused_relu_0 = async_compile.triton('triton_poi_fused_relu_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[32768],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_relu_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_relu_0(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 25600
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 400
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tl.store(in_out_ptr0 + (x2), tmp4, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/37/c37o3rcnefcsc73wadyyeatbmrkuvk5gmf7g2reko6sjih3zzlcv.py
# Topologically Sorted Source Nodes: [mul, std, mul_1, z], Original ATen: [aten.mul, aten.exp, aten.add]
# Source node to ATen node mapping:
# mul => mul
# mul_1 => mul_1
# std => exp
# z => add
# Graph fragment:
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%addmm_2, 0.5), kwargs = {})
# %exp : [num_users=1] = call_function[target=torch.ops.aten.exp.default](args = (%mul,), kwargs = {})
# %mul_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%randn, %exp), kwargs = {})
# %add : [num_users=2] = call_function[target=torch.ops.aten.add.Tensor](args = (%addmm_1, %mul_1), kwargs = {})
triton_poi_fused_add_exp_mul_1 = async_compile.triton('triton_poi_fused_add_exp_mul_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[2048],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_exp_mul_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 3, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_exp_mul_1(in_ptr0, in_ptr1, in_ptr2, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 1280
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (x0), xmask)
tmp1 = tl.load(in_ptr1 + (x0), xmask)
tmp2 = tl.load(in_ptr2 + (x0), xmask)
tmp3 = 0.5
tmp4 = tmp2 * tmp3
tmp5 = tl_math.exp(tmp4)
tmp6 = tmp1 * tmp5
tmp7 = tmp0 + tmp6
tl.store(out_ptr0 + (x0), tmp7, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/ki/ckinq4jozjo44rxyehsa2lhip3xyubjyvbx6rmx44eqvtqzfiasz.py
# Topologically Sorted Source Nodes: [sigmoid], Original ATen: [aten.sigmoid]
# Source node to ATen node mapping:
# sigmoid => sigmoid
# Graph fragment:
# %add_tensor : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mm_default, %primals_11), kwargs = {})
# %sigmoid : [num_users=1] = call_function[target=torch.ops.aten.sigmoid.default](args = (%add_tensor,), kwargs = {})
triton_poi_fused_sigmoid_2 = async_compile.triton('triton_poi_fused_sigmoid_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_sigmoid_2', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_sigmoid_2(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.sigmoid(tmp2)
tl.store(in_out_ptr0 + (x2), tmp3, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9, primals_10, primals_11 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (400, 4), (4, 1))
assert_size_stride(primals_3, (400, ), (1, ))
assert_size_stride(primals_4, (20, 400), (400, 1))
assert_size_stride(primals_5, (20, ), (1, ))
assert_size_stride(primals_6, (20, 400), (400, 1))
assert_size_stride(primals_7, (20, ), (1, ))
assert_size_stride(primals_8, (400, 20), (20, 1))
assert_size_stride(primals_9, (400, ), (1, ))
assert_size_stride(primals_10, (4, 400), (400, 1))
assert_size_stride(primals_11, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 400), (400, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(primals_1, (64, 4), (4, 1), 0), reinterpret_tensor(primals_2, (4, 400), (1, 4), 0), out=buf0)
del primals_2
buf1 = buf0; del buf0 # reuse
# Topologically Sorted Source Nodes: [h1], Original ATen: [aten.relu]
stream0 = get_raw_stream(0)
triton_poi_fused_relu_0.run(buf1, primals_3, 25600, grid=grid(25600), stream=stream0)
del primals_3
buf2 = empty_strided_cuda((64, 20), (20, 1), torch.float32)
# Topologically Sorted Source Nodes: [mu], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_5, buf1, reinterpret_tensor(primals_4, (400, 20), (1, 400), 0), alpha=1, beta=1, out=buf2)
del primals_5
buf3 = empty_strided_cuda((64, 20), (20, 1), torch.float32)
# Topologically Sorted Source Nodes: [logvar], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_7, buf1, reinterpret_tensor(primals_6, (400, 20), (1, 400), 0), alpha=1, beta=1, out=buf3)
del primals_7
# Topologically Sorted Source Nodes: [eps], Original ATen: [aten.randn_like]
buf4 = torch.ops.aten.randn.default([64, 20], dtype=torch.float32, device=device(type='cuda', index=0), pin_memory=False)
buf5 = buf4
del buf4
buf6 = empty_strided_cuda((64, 20), (20, 1), torch.float32)
# Topologically Sorted Source Nodes: [mul, std, mul_1, z], Original ATen: [aten.mul, aten.exp, aten.add]
triton_poi_fused_add_exp_mul_1.run(buf2, buf5, buf3, buf6, 1280, grid=grid(1280), stream=stream0)
buf7 = empty_strided_cuda((64, 400), (400, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(buf6, reinterpret_tensor(primals_8, (20, 400), (1, 20), 0), out=buf7)
buf8 = buf7; del buf7 # reuse
# Topologically Sorted Source Nodes: [h3], Original ATen: [aten.relu]
triton_poi_fused_relu_0.run(buf8, primals_9, 25600, grid=grid(25600), stream=stream0)
del primals_9
buf9 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(buf8, reinterpret_tensor(primals_10, (400, 4), (1, 400), 0), out=buf9)
buf10 = buf9; del buf9 # reuse
# Topologically Sorted Source Nodes: [sigmoid], Original ATen: [aten.sigmoid]
triton_poi_fused_sigmoid_2.run(buf10, primals_11, 256, grid=grid(256), stream=stream0)
del primals_11
return (buf10, buf2, buf3, buf6, reinterpret_tensor(primals_1, (64, 4), (4, 1), 0), buf1, buf3, buf5, buf6, buf8, buf10, primals_10, primals_8, primals_6, primals_4, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((400, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((400, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((20, 400), (400, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((20, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((20, 400), (400, 1), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((20, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_8 = rand_strided((400, 20), (20, 1), device='cuda:0', dtype=torch.float32)
primals_9 = rand_strided((400, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_10 = rand_strided((4, 400), (400, 1), device='cuda:0', dtype=torch.float32)
primals_11 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9, primals_10, primals_11])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.nn.modules.loss
class VAE(nn.Module):
""" Variational Autoencoder for dimensional reduction"""
def __init__(self, dim):
super(VAE, self).__init__()
self.dim = dim
self.fc1 = nn.Linear(dim, 400)
self.fc21 = nn.Linear(400, 20)
self.fc22 = nn.Linear(400, 20)
self.fc3 = nn.Linear(20, 400)
self.fc4 = nn.Linear(400, dim)
def encode(self, x):
h1 = F.relu(self.fc1(x))
return self.fc21(h1), self.fc22(h1)
def reparameterize(self, mu, logvar):
std = torch.exp(0.5 * logvar)
eps = torch.randn_like(std)
return mu + eps * std
def decode(self, z):
h3 = F.relu(self.fc3(z))
return torch.sigmoid(self.fc4(h3))
def forward(self, x):
mu, logvar = self.encode(x.view(-1, self.dim))
z = self.reparameterize(mu, logvar)
return self.decode(z), mu, logvar, z
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'dim': 4}]
| import torch
from torch import device
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import math as tl_math
import torch.nn as nn
import torch.nn.functional as F
import torch.nn.modules.loss
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_relu_0(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 25600
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 400
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tl.store(in_out_ptr0 + x2, tmp4, xmask)
@triton.jit
def triton_poi_fused_add_exp_mul_1(in_ptr0, in_ptr1, in_ptr2, out_ptr0,
xnumel, XBLOCK: tl.constexpr):
xnumel = 1280
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + x0, xmask)
tmp1 = tl.load(in_ptr1 + x0, xmask)
tmp2 = tl.load(in_ptr2 + x0, xmask)
tmp3 = 0.5
tmp4 = tmp2 * tmp3
tmp5 = tl_math.exp(tmp4)
tmp6 = tmp1 * tmp5
tmp7 = tmp0 + tmp6
tl.store(out_ptr0 + x0, tmp7, xmask)
@triton.jit
def triton_poi_fused_sigmoid_2(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl.
constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.sigmoid(tmp2)
tl.store(in_out_ptr0 + x2, tmp3, xmask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7, primals_8, primals_9, primals_10, primals_11) = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (400, 4), (4, 1))
assert_size_stride(primals_3, (400,), (1,))
assert_size_stride(primals_4, (20, 400), (400, 1))
assert_size_stride(primals_5, (20,), (1,))
assert_size_stride(primals_6, (20, 400), (400, 1))
assert_size_stride(primals_7, (20,), (1,))
assert_size_stride(primals_8, (400, 20), (20, 1))
assert_size_stride(primals_9, (400,), (1,))
assert_size_stride(primals_10, (4, 400), (400, 1))
assert_size_stride(primals_11, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 400), (400, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_1, (64, 4), (4, 1), 0),
reinterpret_tensor(primals_2, (4, 400), (1, 4), 0), out=buf0)
del primals_2
buf1 = buf0
del buf0
get_raw_stream(0)
triton_poi_fused_relu_0[grid(25600)](buf1, primals_3, 25600, XBLOCK
=128, num_warps=4, num_stages=1)
del primals_3
buf2 = empty_strided_cuda((64, 20), (20, 1), torch.float32)
extern_kernels.addmm(primals_5, buf1, reinterpret_tensor(primals_4,
(400, 20), (1, 400), 0), alpha=1, beta=1, out=buf2)
del primals_5
buf3 = empty_strided_cuda((64, 20), (20, 1), torch.float32)
extern_kernels.addmm(primals_7, buf1, reinterpret_tensor(primals_6,
(400, 20), (1, 400), 0), alpha=1, beta=1, out=buf3)
del primals_7
buf4 = torch.ops.aten.randn.default([64, 20], dtype=torch.float32,
device=device(type='cuda', index=0), pin_memory=False)
buf5 = buf4
del buf4
buf6 = empty_strided_cuda((64, 20), (20, 1), torch.float32)
triton_poi_fused_add_exp_mul_1[grid(1280)](buf2, buf5, buf3, buf6,
1280, XBLOCK=128, num_warps=4, num_stages=1)
buf7 = empty_strided_cuda((64, 400), (400, 1), torch.float32)
extern_kernels.mm(buf6, reinterpret_tensor(primals_8, (20, 400), (1,
20), 0), out=buf7)
buf8 = buf7
del buf7
triton_poi_fused_relu_0[grid(25600)](buf8, primals_9, 25600, XBLOCK
=128, num_warps=4, num_stages=1)
del primals_9
buf9 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
extern_kernels.mm(buf8, reinterpret_tensor(primals_10, (400, 4), (1,
400), 0), out=buf9)
buf10 = buf9
del buf9
triton_poi_fused_sigmoid_2[grid(256)](buf10, primals_11, 256,
XBLOCK=256, num_warps=4, num_stages=1)
del primals_11
return (buf10, buf2, buf3, buf6, reinterpret_tensor(primals_1, (64, 4),
(4, 1), 0), buf1, buf3, buf5, buf6, buf8, buf10, primals_10,
primals_8, primals_6, primals_4)
class VAENew(nn.Module):
""" Variational Autoencoder for dimensional reduction"""
def __init__(self, dim):
super(VAENew, self).__init__()
self.dim = dim
self.fc1 = nn.Linear(dim, 400)
self.fc21 = nn.Linear(400, 20)
self.fc22 = nn.Linear(400, 20)
self.fc3 = nn.Linear(20, 400)
self.fc4 = nn.Linear(400, dim)
def encode(self, x):
h1 = F.relu(self.fc1(x))
return self.fc21(h1), self.fc22(h1)
def reparameterize(self, mu, logvar):
std = torch.exp(0.5 * logvar)
eps = torch.randn_like(std)
return mu + eps * std
def decode(self, z):
h3 = F.relu(self.fc3(z))
return torch.sigmoid(self.fc4(h3))
def forward(self, input_0):
primals_2 = self.fc1.weight
primals_3 = self.fc1.bias
primals_4 = self.fc21.weight
primals_5 = self.fc21.bias
primals_6 = self.fc22.weight
primals_7 = self.fc22.bias
primals_8 = self.fc3.weight
primals_9 = self.fc3.bias
primals_10 = self.fc4.weight
primals_11 = self.fc4.bias
primals_1 = input_0
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7, primals_8, primals_9,
primals_10, primals_11])
return output[0], output[1], output[2], output[3]
| hurraygong/scGNN | VAE | false | 6,835 | [
"MIT"
] | 1 | bc555895fbd5740ddd82e03187171116889cc10e | https://github.com/hurraygong/scGNN/tree/bc555895fbd5740ddd82e03187171116889cc10e | import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.nn.modules.loss
class Model(nn.Module):
""" Variational Autoencoder for dimensional reduction"""
def __init__(self, dim):
super().__init__()
self.dim = dim
self.fc1 = nn.Linear(dim, 400)
self.fc21 = nn.Linear(400, 20)
self.fc22 = nn.Linear(400, 20)
self.fc3 = nn.Linear(20, 400)
self.fc4 = nn.Linear(400, dim)
def encode(self, x):
h1 = F.relu(self.fc1(x))
return self.fc21(h1), self.fc22(h1)
def reparameterize(self, mu, logvar):
std = torch.exp(0.5 * logvar)
eps = torch.randn_like(std)
return mu + eps * std
def decode(self, z):
h3 = F.relu(self.fc3(z))
return torch.sigmoid(self.fc4(h3))
def forward(self, x):
mu, logvar = self.encode(x.view(-1, self.dim))
z = self.reparameterize(mu, logvar)
return self.decode(z), mu, logvar, z
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4]
|
eca_block | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/l3/cl35tzbhrd24dhunkbb6gjs54aklpyr46oikqhoylcgmkcmhujil.py
# Topologically Sorted Source Nodes: [y], Original ATen: [aten.mean]
# Source node to ATen node mapping:
# y => mean
# Graph fragment:
# %mean : [num_users=1] = call_function[target=torch.ops.aten.mean.dim](args = (%primals_1, [-1, -2], True), kwargs = {})
triton_per_fused_mean_0 = async_compile.triton('triton_per_fused_mean_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[16, 16],
reduction_hint=ReductionHint.INNER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused_mean_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 1, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused_mean_0(in_out_ptr0, in_ptr0, xnumel, rnumel, XBLOCK : tl.constexpr):
xnumel = 16
rnumel = 16
RBLOCK: tl.constexpr = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
roffset = 0
rmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
r1 = rindex
x0 = xindex
tmp0 = tl.load(in_ptr0 + (r1 + (16*x0)), xmask, other=0.0)
tmp1 = tl.broadcast_to(tmp0, [XBLOCK, RBLOCK])
tmp3 = tl.where(xmask, tmp1, 0)
tmp4 = tl.sum(tmp3, 1)[:, None]
tmp5 = 16.0
tmp6 = tmp4 / tmp5
tl.debug_barrier()
tl.store(in_out_ptr0 + (x0), tmp6, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/hx/chxvgixiwduvwuumo7j2hhpjfvzwfh7g2wp26wd4453y6egzxpmt.py
# Topologically Sorted Source Nodes: [mul], Original ATen: [aten.mul]
# Source node to ATen node mapping:
# mul => mul
# Graph fragment:
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%primals_1, %expand), kwargs = {})
triton_poi_fused_mul_1 = async_compile.triton('triton_poi_fused_mul_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_mul_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_mul_1(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 16)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr1 + (x1), xmask, eviction_policy='evict_last')
tmp2 = tl.sigmoid(tmp1)
tmp3 = tmp0 * tmp2
tl.store(out_ptr0 + (x2), tmp3, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (1, 1, 1), (1, 1, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 1, 1), (4, 1, 16, 16), torch.float32)
buf1 = buf0; del buf0 # reuse
# Topologically Sorted Source Nodes: [y], Original ATen: [aten.mean]
stream0 = get_raw_stream(0)
triton_per_fused_mean_0.run(buf1, primals_1, 16, 16, grid=grid(16), stream=stream0)
# Topologically Sorted Source Nodes: [conv1d], Original ATen: [aten.convolution]
buf2 = extern_kernels.convolution(reinterpret_tensor(buf1, (4, 1, 4), (4, 0, 1), 0), primals_2, stride=(1,), padding=(0,), dilation=(1,), transposed=False, output_padding=(0,), groups=1, bias=None)
assert_size_stride(buf2, (4, 1, 4), (4, 4, 1))
buf3 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [mul], Original ATen: [aten.mul]
triton_poi_fused_mul_1.run(primals_1, buf2, buf3, 256, grid=grid(256), stream=stream0)
return (buf3, primals_1, primals_2, reinterpret_tensor(buf1, (4, 1, 4), (4, 1, 1), 0), buf2, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((1, 1, 1), (1, 1, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import math
import torch
import torch.nn as nn
class eca_block(nn.Module):
def __init__(self, channel, b=1, gamma=2):
super(eca_block, self).__init__()
kernel_size = int(abs((math.log(channel, 2) + b) / gamma))
kernel_size = kernel_size if kernel_size % 2 else kernel_size + 1
self.avg_pool = nn.AdaptiveAvgPool2d(1)
self.conv = nn.Conv1d(1, 1, kernel_size=kernel_size, padding=(
kernel_size - 1) // 2, bias=False)
self.sigmoid = nn.Sigmoid()
def forward(self, x):
y = self.avg_pool(x)
y = self.conv(y.squeeze(-1).transpose(-1, -2)).transpose(-1, -2
).unsqueeze(-1)
y = self.sigmoid(y)
return x * y.expand_as(x)
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'channel': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
import math
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_per_fused_mean_0(in_out_ptr0, in_ptr0, xnumel, rnumel, XBLOCK:
tl.constexpr):
xnumel = 16
RBLOCK: tl.constexpr = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
r1 = rindex
x0 = xindex
tmp0 = tl.load(in_ptr0 + (r1 + 16 * x0), xmask, other=0.0)
tmp1 = tl.broadcast_to(tmp0, [XBLOCK, RBLOCK])
tmp3 = tl.where(xmask, tmp1, 0)
tmp4 = tl.sum(tmp3, 1)[:, None]
tmp5 = 16.0
tmp6 = tmp4 / tmp5
tl.debug_barrier()
tl.store(in_out_ptr0 + x0, tmp6, xmask)
@triton.jit
def triton_poi_fused_mul_1(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK: tl.
constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 16
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr1 + x1, xmask, eviction_policy='evict_last')
tmp2 = tl.sigmoid(tmp1)
tmp3 = tmp0 * tmp2
tl.store(out_ptr0 + x2, tmp3, xmask)
def call(args):
primals_1, primals_2 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (1, 1, 1), (1, 1, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 1, 1), (4, 1, 16, 16), torch.float32)
buf1 = buf0
del buf0
get_raw_stream(0)
triton_per_fused_mean_0[grid(16)](buf1, primals_1, 16, 16, XBLOCK=8,
num_warps=2, num_stages=1)
buf2 = extern_kernels.convolution(reinterpret_tensor(buf1, (4, 1, 4
), (4, 0, 1), 0), primals_2, stride=(1,), padding=(0,),
dilation=(1,), transposed=False, output_padding=(0,), groups=1,
bias=None)
assert_size_stride(buf2, (4, 1, 4), (4, 4, 1))
buf3 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
triton_poi_fused_mul_1[grid(256)](primals_1, buf2, buf3, 256,
XBLOCK=256, num_warps=4, num_stages=1)
return buf3, primals_1, primals_2, reinterpret_tensor(buf1, (4, 1, 4),
(4, 1, 1), 0), buf2
class eca_blockNew(nn.Module):
def __init__(self, channel, b=1, gamma=2):
super(eca_blockNew, self).__init__()
kernel_size = int(abs((math.log(channel, 2) + b) / gamma))
kernel_size = kernel_size if kernel_size % 2 else kernel_size + 1
self.avg_pool = nn.AdaptiveAvgPool2d(1)
self.conv = nn.Conv1d(1, 1, kernel_size=kernel_size, padding=(
kernel_size - 1) // 2, bias=False)
self.sigmoid = nn.Sigmoid()
def forward(self, input_0):
primals_2 = self.conv.weight
primals_1 = input_0
output = call([primals_1, primals_2])
return output[0]
| huuthieu/pytorch-yolov4-tiny | eca_block | false | 6,836 | [
"MIT"
] | 1 | fac82da75e161221af74b56242272a42cf64c17e | https://github.com/huuthieu/pytorch-yolov4-tiny/tree/fac82da75e161221af74b56242272a42cf64c17e | import math
import torch
import torch.nn as nn
class Model(nn.Module):
def __init__(self, channel, b=1, gamma=2):
super().__init__()
kernel_size = int(abs((math.log(channel, 2) + b) / gamma))
kernel_size = kernel_size if kernel_size % 2 else kernel_size + 1
self.avg_pool = nn.AdaptiveAvgPool2d(1)
self.conv = nn.Conv1d(1, 1, kernel_size=kernel_size, padding=(
kernel_size - 1) // 2, bias=False)
self.sigmoid = nn.Sigmoid()
def forward(self, x):
y = self.avg_pool(x)
y = self.conv(y.squeeze(-1).transpose(-1, -2)).transpose(-1, -2
).unsqueeze(-1)
y = self.sigmoid(y)
return x * y.expand_as(x)
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4]
|
IrisClassifier | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/nu/cnuuaznpt4szfn74bn46qfjkdypvlkfa5x44ywjpperdjt2a66rj.py
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.relu, aten.threshold_backward]
# Source node to ATen node mapping:
# x => relu
# Graph fragment:
# %relu : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%view_1,), kwargs = {})
# %le_1 : [num_users=1] = call_function[target=torch.ops.aten.le.Scalar](args = (%relu, 0), kwargs = {})
triton_poi_fused_relu_threshold_backward_0 = async_compile.triton('triton_poi_fused_relu_threshold_backward_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[1024],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*i1', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_relu_threshold_backward_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_relu_threshold_backward_0(in_out_ptr0, in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 640
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 10
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(in_out_ptr0 + (x2), tmp4, xmask)
tl.store(out_ptr0 + (x2), tmp6, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7 = args
args.clear()
assert_size_stride(primals_1, (10, 4), (4, 1))
assert_size_stride(primals_2, (10, ), (1, ))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_4, (10, 10), (10, 1))
assert_size_stride(primals_5, (10, ), (1, ))
assert_size_stride(primals_6, (3, 10), (10, 1))
assert_size_stride(primals_7, (3, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 10), (10, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(primals_3, (64, 4), (4, 1), 0), reinterpret_tensor(primals_1, (4, 10), (1, 4), 0), out=buf0)
del primals_1
buf1 = reinterpret_tensor(buf0, (4, 4, 4, 10), (160, 40, 10, 1), 0); del buf0 # reuse
buf9 = empty_strided_cuda((4, 4, 4, 10), (160, 40, 10, 1), torch.bool)
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.relu, aten.threshold_backward]
stream0 = get_raw_stream(0)
triton_poi_fused_relu_threshold_backward_0.run(buf1, primals_2, buf9, 640, grid=grid(640), stream=stream0)
del primals_2
buf2 = empty_strided_cuda((64, 10), (10, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(buf1, (64, 10), (10, 1), 0), reinterpret_tensor(primals_4, (10, 10), (1, 10), 0), out=buf2)
buf3 = reinterpret_tensor(buf2, (4, 4, 4, 10), (160, 40, 10, 1), 0); del buf2 # reuse
buf8 = empty_strided_cuda((4, 4, 4, 10), (160, 40, 10, 1), torch.bool)
# Topologically Sorted Source Nodes: [x_1], Original ATen: [aten.relu, aten.threshold_backward]
triton_poi_fused_relu_threshold_backward_0.run(buf3, primals_5, buf8, 640, grid=grid(640), stream=stream0)
del primals_5
# Topologically Sorted Source Nodes: [x_1, x_2], Original ATen: [aten.relu, aten.native_dropout]
buf4 = torch.ops.aten.native_dropout.default(buf3, 0.2, True)
del buf3
buf5 = buf4[0]
buf6 = buf4[1]
del buf4
buf7 = empty_strided_cuda((64, 3), (3, 1), torch.float32)
# Topologically Sorted Source Nodes: [x_3], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_7, reinterpret_tensor(buf5, (64, 10), (10, 1), 0), reinterpret_tensor(primals_6, (10, 3), (1, 10), 0), alpha=1, beta=1, out=buf7)
del primals_7
return (reinterpret_tensor(buf7, (4, 4, 4, 3), (48, 12, 3, 1), 0), reinterpret_tensor(primals_3, (64, 4), (4, 1), 0), reinterpret_tensor(buf1, (64, 10), (10, 1), 0), buf6, reinterpret_tensor(buf5, (64, 10), (10, 1), 0), primals_6, buf8, primals_4, buf9, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((10, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((10, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((10, 10), (10, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((10, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((3, 10), (10, 1), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((3, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.onnx
class IrisClassifier(nn.Module):
def __init__(self):
super(IrisClassifier, self).__init__()
self.fc1 = nn.Linear(4, 10)
self.fc2 = nn.Linear(10, 10)
self.fc3 = nn.Linear(10, 3)
def forward(self, x):
x = F.relu(self.fc1(x))
x = F.relu(self.fc2(x))
x = F.dropout(x, 0.2)
x = self.fc3(x)
return x
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
import torch.nn as nn
import torch.onnx
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_relu_threshold_backward_0(in_out_ptr0, in_ptr0,
out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 640
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 10
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(in_out_ptr0 + x2, tmp4, xmask)
tl.store(out_ptr0 + x2, tmp6, xmask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7) = args
args.clear()
assert_size_stride(primals_1, (10, 4), (4, 1))
assert_size_stride(primals_2, (10,), (1,))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_4, (10, 10), (10, 1))
assert_size_stride(primals_5, (10,), (1,))
assert_size_stride(primals_6, (3, 10), (10, 1))
assert_size_stride(primals_7, (3,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 10), (10, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_3, (64, 4), (4, 1), 0),
reinterpret_tensor(primals_1, (4, 10), (1, 4), 0), out=buf0)
del primals_1
buf1 = reinterpret_tensor(buf0, (4, 4, 4, 10), (160, 40, 10, 1), 0)
del buf0
buf9 = empty_strided_cuda((4, 4, 4, 10), (160, 40, 10, 1), torch.bool)
get_raw_stream(0)
triton_poi_fused_relu_threshold_backward_0[grid(640)](buf1,
primals_2, buf9, 640, XBLOCK=128, num_warps=4, num_stages=1)
del primals_2
buf2 = empty_strided_cuda((64, 10), (10, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(buf1, (64, 10), (10, 1), 0),
reinterpret_tensor(primals_4, (10, 10), (1, 10), 0), out=buf2)
buf3 = reinterpret_tensor(buf2, (4, 4, 4, 10), (160, 40, 10, 1), 0)
del buf2
buf8 = empty_strided_cuda((4, 4, 4, 10), (160, 40, 10, 1), torch.bool)
triton_poi_fused_relu_threshold_backward_0[grid(640)](buf3,
primals_5, buf8, 640, XBLOCK=128, num_warps=4, num_stages=1)
del primals_5
buf4 = torch.ops.aten.native_dropout.default(buf3, 0.2, True)
del buf3
buf5 = buf4[0]
buf6 = buf4[1]
del buf4
buf7 = empty_strided_cuda((64, 3), (3, 1), torch.float32)
extern_kernels.addmm(primals_7, reinterpret_tensor(buf5, (64, 10),
(10, 1), 0), reinterpret_tensor(primals_6, (10, 3), (1, 10), 0),
alpha=1, beta=1, out=buf7)
del primals_7
return reinterpret_tensor(buf7, (4, 4, 4, 3), (48, 12, 3, 1), 0
), reinterpret_tensor(primals_3, (64, 4), (4, 1), 0
), reinterpret_tensor(buf1, (64, 10), (10, 1), 0
), buf6, reinterpret_tensor(buf5, (64, 10), (10, 1), 0
), primals_6, buf8, primals_4, buf9
class IrisClassifierNew(nn.Module):
def __init__(self):
super(IrisClassifierNew, self).__init__()
self.fc1 = nn.Linear(4, 10)
self.fc2 = nn.Linear(10, 10)
self.fc3 = nn.Linear(10, 3)
def forward(self, input_0):
primals_1 = self.fc1.weight
primals_2 = self.fc1.bias
primals_4 = self.fc2.weight
primals_5 = self.fc2.bias
primals_6 = self.fc3.weight
primals_7 = self.fc3.bias
primals_3 = input_0
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7])
return output[0]
| huxuan/mlflow | IrisClassifier | false | 6,837 | [
"Apache-2.0"
] | 1 | 7b4ab0e4cac5d4c2d2cbfcd3d12aa55b2ee83efe | https://github.com/huxuan/mlflow/tree/7b4ab0e4cac5d4c2d2cbfcd3d12aa55b2ee83efe | import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.onnx
class Model(nn.Module):
def __init__(self):
super().__init__()
self.fc1 = nn.Linear(4, 10)
self.fc2 = nn.Linear(10, 10)
self.fc3 = nn.Linear(10, 3)
def forward(self, x):
x = F.relu(self.fc1(x))
x = F.relu(self.fc2(x))
x = F.dropout(x, 0.2)
x = self.fc3(x)
return x
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return []
|
DynamicPreHead | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/kn/cknyjwkwufnzzf4ya3scui55ownkmt5cdh3hggzwsfe3ch5fshzm.py
# Unsorted Source Nodes: [], Original ATen: []
# Source node to ATen node mapping:
triton_poi_fused_0 = async_compile.triton('triton_poi_fused_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16, 4096], tile_hint=TileHint.SQUARE,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_0(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 12
xnumel = 4096
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = tl.full([XBLOCK, YBLOCK], True, tl.int1)
x2 = xindex
y3 = yindex
y0 = yindex % 3
y1 = (yindex // 3)
tmp0 = tl.load(in_ptr0 + (x2 + (4096*y3)), ymask, eviction_policy='evict_last')
tl.store(out_ptr0 + (y0 + (3*x2) + (12288*y1)), tmp0, ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/5n/c5ntvxirzrfxhvpqxhnwygvq7ljms62jucexb6q4gzohhvbidcan.py
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.convolution]
# Source node to ATen node mapping:
# x => convolution
# Graph fragment:
# %convolution : [num_users=2] = call_function[target=torch.ops.aten.convolution.default](args = (%primals_3, %primals_1, %primals_2, [1, 1], [0, 0], [1, 1], False, [0, 0], 1), kwargs = {})
triton_poi_fused_convolution_1 = async_compile.triton('triton_poi_fused_convolution_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[2097152],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_1', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_1(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 1638400
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = tl.full([XBLOCK], True, tl.int1)
x2 = xindex
x0 = xindex % 100
tmp0 = tl.load(in_out_ptr0 + (x2), None)
tmp1 = tl.load(in_ptr0 + (x0), None, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + (x2), tmp2, None)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/tc/ctc2mg6575smgmi5ym7g47qyuwaxds2hntyzthl2dmq46vajn25m.py
# Topologically Sorted Source Nodes: [x_1], Original ATen: [aten.native_group_norm]
# Source node to ATen node mapping:
# x_1 => var_mean
# Graph fragment:
# %var_mean : [num_users=2] = call_function[target=torch.ops.aten.var_mean.correction](args = (%view, [2, 3]), kwargs = {correction: 0, keepdim: True})
triton_per_fused_native_group_norm_2 = async_compile.triton('triton_per_fused_native_group_norm_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[16384, 128],
reduction_hint=ReductionHint.INNER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32', 5: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused_native_group_norm_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 5, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused_native_group_norm_2(in_ptr0, out_ptr0, out_ptr1, out_ptr2, xnumel, rnumel, XBLOCK : tl.constexpr):
xnumel = 12800
rnumel = 128
RBLOCK: tl.constexpr = 128
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
roffset = 0
rmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
r3 = rindex
x0 = xindex % 50
x1 = (xindex // 50) % 64
x2 = (xindex // 3200)
x4 = xindex
tmp0 = tl.load(in_ptr0 + ((2*x0) + (100*((r3 + (128*x1)) % 4096)) + (409600*x2) + ((r3 + (128*x1)) // 4096)), xmask, eviction_policy='evict_last', other=0.0)
tmp1 = tl.broadcast_to(tmp0, [XBLOCK, RBLOCK])
tmp3 = tl.where(xmask, tmp1, 0)
tmp4 = tl.broadcast_to(tmp1, [XBLOCK, RBLOCK])
tmp6 = tl.where(xmask, tmp4, 0)
tmp7 = tl.sum(tmp6, 1)[:, None]
tmp8 = tl.full([XBLOCK, 1], 128, tl.int32)
tmp9 = tmp8.to(tl.float32)
tmp10 = tmp7 / tmp9
tmp11 = tmp1 - tmp10
tmp12 = tmp11 * tmp11
tmp13 = tl.broadcast_to(tmp12, [XBLOCK, RBLOCK])
tmp15 = tl.where(xmask, tmp13, 0)
tmp16 = tl.sum(tmp15, 1)[:, None]
tl.store(out_ptr0 + (x4), tmp10, xmask)
tl.store(out_ptr1 + (x4), tmp16, xmask)
tl.store(out_ptr2 + (x4), tmp9, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/r7/cr7y4cpmphvbdki47z6sjwhqvpap6fmimynrcbmpoclzdojb5elw.py
# Topologically Sorted Source Nodes: [x_1], Original ATen: [aten.native_group_norm]
# Source node to ATen node mapping:
# x_1 => var_mean
# Graph fragment:
# %var_mean : [num_users=2] = call_function[target=torch.ops.aten.var_mean.correction](args = (%view, [2, 3]), kwargs = {correction: 0, keepdim: True})
triton_per_fused_native_group_norm_3 = async_compile.triton('triton_per_fused_native_group_norm_3', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[256, 64],
reduction_hint=ReductionHint.INNER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: '*fp32', 5: '*fp32', 6: 'i32', 7: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5, 7), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused_native_group_norm_3', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 3, 'num_reduction': 3, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused_native_group_norm_3(in_ptr0, in_ptr1, in_ptr2, out_ptr0, out_ptr1, out_ptr2, xnumel, rnumel, XBLOCK : tl.constexpr):
xnumel = 200
rnumel = 64
RBLOCK: tl.constexpr = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
roffset = 0
rmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
r2 = rindex
x0 = xindex % 50
x1 = (xindex // 50)
x3 = xindex
tmp0 = tl.load(in_ptr0 + (x0 + (50*r2) + (3200*x1)), xmask, other=0.0)
tmp1 = tl.load(in_ptr1 + (x0 + (50*r2) + (3200*x1)), xmask, other=0.0)
tmp2 = tl.load(in_ptr2 + (x0 + (50*r2) + (3200*x1)), xmask, other=0.0)
tmp3 = tl.broadcast_to(tmp0, [XBLOCK, RBLOCK])
tmp4 = tl.broadcast_to(tmp1, [XBLOCK, RBLOCK])
tmp5 = tl.broadcast_to(tmp2, [XBLOCK, RBLOCK])
tmp7 = tl.where(xmask, tmp3, 0)
tmp8 = tl.where(xmask, tmp4, 0)
tmp9 = tl.where(xmask, tmp5, 0)
tmp10, tmp11, tmp12 = triton_helpers.welford(tmp7, tmp8, tmp9, 1)
tmp13 = tmp10[:, None]
tmp14 = tmp11[:, None]
tmp15 = tmp12[:, None]
tl.store(out_ptr0 + (x3), tmp13, xmask)
tl.store(out_ptr1 + (x3), tmp14, xmask)
tl.store(out_ptr2 + (x3), tmp15, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/gd/cgdldbamtse56reif25lhp56oz47dwxhzalv2yrz7d2kd7wnokg3.py
# Topologically Sorted Source Nodes: [x_1], Original ATen: [aten.native_group_norm]
# Source node to ATen node mapping:
# x_1 => add, rsqrt, var_mean
# Graph fragment:
# %var_mean : [num_users=2] = call_function[target=torch.ops.aten.var_mean.correction](args = (%view, [2, 3]), kwargs = {correction: 0, keepdim: True})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%getitem, 1e-05), kwargs = {})
# %rsqrt : [num_users=2] = call_function[target=torch.ops.aten.rsqrt.default](args = (%add,), kwargs = {})
triton_per_fused_native_group_norm_4 = async_compile.triton('triton_per_fused_native_group_norm_4', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[128, 2],
reduction_hint=ReductionHint.INNER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: '*fp32', 5: '*fp32', 6: 'i32', 7: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused_native_group_norm_4', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 3, 'num_reduction': 2, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused_native_group_norm_4(in_ptr0, in_ptr1, in_ptr2, out_ptr0, out_ptr1, out_ptr2, xnumel, rnumel, XBLOCK : tl.constexpr):
xnumel = 100
rnumel = 2
RBLOCK: tl.constexpr = 2
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
roffset = 0
rmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
r1 = rindex
x0 = xindex
tmp0 = tl.load(in_ptr0 + (r1 + (2*x0)), xmask, other=0.0)
tmp1 = tl.load(in_ptr1 + (r1 + (2*x0)), xmask, other=0.0)
tmp2 = tl.load(in_ptr2 + (r1 + (2*x0)), xmask, other=0.0)
tmp3 = tl.broadcast_to(tmp0, [XBLOCK, RBLOCK])
tmp4 = tl.broadcast_to(tmp1, [XBLOCK, RBLOCK])
tmp5 = tl.broadcast_to(tmp2, [XBLOCK, RBLOCK])
tmp7 = tl.where(xmask, tmp3, 0)
tmp8 = tl.where(xmask, tmp4, 0)
tmp9 = tl.where(xmask, tmp5, 0)
tmp10, tmp11, tmp12 = triton_helpers.welford(tmp7, tmp8, tmp9, 1)
tmp13 = tmp10[:, None]
tmp14 = tmp11[:, None]
tmp15 = tmp12[:, None]
tmp16 = 16384.0
tmp17 = tmp14 / tmp16
tmp18 = 1e-05
tmp19 = tmp17 + tmp18
tmp20 = libdevice.rsqrt(tmp19)
tl.store(out_ptr2 + (x0), tmp20, xmask)
tl.store(out_ptr0 + (x0), tmp13, xmask)
tl.store(out_ptr1 + (x0), tmp14, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/o5/co55ow27kxqru5gjpedo4tmhg6oyewjwn2aduyikuiqnn3wmpzwf.py
# Topologically Sorted Source Nodes: [x_1, x_2], Original ATen: [aten.native_group_norm, aten.relu, aten.threshold_backward]
# Source node to ATen node mapping:
# x_1 => add_1, mul_1
# x_2 => relu
# Graph fragment:
# %mul_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%view_1, %unsqueeze_5), kwargs = {})
# %add_1 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul_1, %unsqueeze_2), kwargs = {})
# %relu : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%add_1,), kwargs = {})
# %le : [num_users=1] = call_function[target=torch.ops.aten.le.Scalar](args = (%relu, 0), kwargs = {})
triton_poi_fused_native_group_norm_relu_threshold_backward_5 = async_compile.triton('triton_poi_fused_native_group_norm_relu_threshold_backward_5', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[512, 4096], tile_hint=TileHint.DEFAULT,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: '*fp32', 5: '*fp32', 6: '*i1', 7: 'i32', 8: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5, 6, 7, 8), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_native_group_norm_relu_threshold_backward_5', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_native_group_norm_relu_threshold_backward_5(in_ptr0, in_ptr1, in_ptr2, in_ptr3, in_ptr4, out_ptr0, out_ptr1, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 400
xnumel = 4096
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = tl.full([XBLOCK, YBLOCK], True, tl.int1)
x2 = xindex
y0 = yindex % 100
y1 = (yindex // 100)
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + (100*x2) + (409600*y1)), ymask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + ((y3 // 4)), ymask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr2 + ((y3 // 4)), ymask, eviction_policy='evict_last')
tmp10 = tl.load(in_ptr3 + (y0), ymask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr4 + (y0), ymask, eviction_policy='evict_last')
tmp2 = tmp0 - tmp1
tmp4 = 16384.0
tmp5 = tmp3 / tmp4
tmp6 = 1e-05
tmp7 = tmp5 + tmp6
tmp8 = libdevice.rsqrt(tmp7)
tmp9 = tmp2 * tmp8
tmp11 = tmp9 * tmp10
tmp13 = tmp11 + tmp12
tmp14 = tl.full([1, 1], 0, tl.int32)
tmp15 = triton_helpers.maximum(tmp14, tmp13)
tmp16 = 0.0
tmp17 = tmp15 <= tmp16
tl.store(out_ptr0 + (x2 + (4096*y3)), tmp15, ymask)
tl.store(out_ptr1 + (y0 + (100*x2) + (409600*y1)), tmp17, ymask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5 = args
args.clear()
assert_size_stride(primals_1, (100, 3, 1, 1), (3, 1, 1, 1))
assert_size_stride(primals_2, (100, ), (1, ))
assert_size_stride(primals_3, (4, 3, 64, 64), (12288, 4096, 64, 1))
assert_size_stride(primals_4, (100, ), (1, ))
assert_size_stride(primals_5, (100, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 3, 64, 64), (12288, 1, 192, 3), torch.float32)
# Unsorted Source Nodes: [], Original ATen: []
stream0 = get_raw_stream(0)
triton_poi_fused_0.run(primals_3, buf0, 12, 4096, grid=grid(12, 4096), stream=stream0)
del primals_3
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.convolution]
buf1 = extern_kernels.convolution(buf0, primals_1, stride=(1, 1), padding=(0, 0), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf1, (4, 100, 64, 64), (409600, 1, 6400, 100))
buf2 = buf1; del buf1 # reuse
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.convolution]
triton_poi_fused_convolution_1.run(buf2, primals_2, 1638400, grid=grid(1638400), stream=stream0)
del primals_2
buf3 = empty_strided_cuda((4, 25, 1, 1, 2, 64), (3200, 2, 12800, 12800, 1, 50), torch.float32)
buf4 = empty_strided_cuda((4, 25, 1, 1, 2, 64), (3200, 2, 12800, 12800, 1, 50), torch.float32)
buf5 = empty_strided_cuda((4, 25, 1, 1, 2, 64), (3200, 2, 12800, 12800, 1, 50), torch.float32)
# Topologically Sorted Source Nodes: [x_1], Original ATen: [aten.native_group_norm]
triton_per_fused_native_group_norm_2.run(buf2, buf3, buf4, buf5, 12800, 128, grid=grid(12800), stream=stream0)
buf6 = empty_strided_cuda((4, 25, 1, 1, 2), (50, 2, 200, 200, 1), torch.float32)
buf7 = empty_strided_cuda((4, 25, 1, 1, 2), (50, 2, 200, 200, 1), torch.float32)
buf8 = empty_strided_cuda((4, 25, 1, 1, 2), (50, 2, 200, 200, 1), torch.float32)
# Topologically Sorted Source Nodes: [x_1], Original ATen: [aten.native_group_norm]
triton_per_fused_native_group_norm_3.run(buf3, buf4, buf5, buf6, buf7, buf8, 200, 64, grid=grid(200), stream=stream0)
del buf3
del buf4
del buf5
buf9 = empty_strided_cuda((4, 25, 1, 1), (25, 1, 100, 100), torch.float32)
buf10 = empty_strided_cuda((4, 25, 1, 1), (25, 1, 100, 100), torch.float32)
buf12 = empty_strided_cuda((4, 25, 1, 1), (25, 1, 100, 100), torch.float32)
# Topologically Sorted Source Nodes: [x_1], Original ATen: [aten.native_group_norm]
triton_per_fused_native_group_norm_4.run(buf6, buf7, buf8, buf9, buf10, buf12, 100, 2, grid=grid(100), stream=stream0)
del buf6
del buf7
del buf8
buf13 = empty_strided_cuda((4, 100, 64, 64), (409600, 4096, 64, 1), torch.float32)
buf14 = empty_strided_cuda((4, 100, 64, 64), (409600, 1, 6400, 100), torch.bool)
# Topologically Sorted Source Nodes: [x_1, x_2], Original ATen: [aten.native_group_norm, aten.relu, aten.threshold_backward]
triton_poi_fused_native_group_norm_relu_threshold_backward_5.run(buf2, buf9, buf10, primals_4, primals_5, buf13, buf14, 400, 4096, grid=grid(400, 4096), stream=stream0)
del buf10
del primals_5
return (buf13, primals_1, buf0, primals_4, buf2, reinterpret_tensor(buf9, (4, 25), (25, 1), 0), reinterpret_tensor(buf12, (4, 25), (25, 1), 0), buf14, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((100, 3, 1, 1), (3, 1, 1, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((100, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 3, 64, 64), (12288, 4096, 64, 1), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((100, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((100, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
class DynamicPreHead(nn.Module):
def __init__(self, in_dim=3, embed_dim=100, kernel_size=1):
super(DynamicPreHead, self).__init__()
self.conv = nn.Conv2d(in_dim, embed_dim, kernel_size=kernel_size,
stride=1, padding=int((kernel_size - 1) / 2))
self.bn = nn.GroupNorm(int(embed_dim / 4), embed_dim)
self.relu = nn.ReLU(True)
nn.init.kaiming_normal_(self.conv.weight, mode='fan_out',
nonlinearity='relu')
def forward(self, x):
x = self.conv(x)
x = self.bn(x)
x = self.relu(x)
return x
def get_inputs():
return [torch.rand([4, 3, 64, 64])]
def get_init_inputs():
return [[], {}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import libdevice
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_0(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK: tl.
constexpr, XBLOCK: tl.constexpr):
ynumel = 12
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
tl.full([XBLOCK, YBLOCK], True, tl.int1)
x2 = xindex
y3 = yindex
y0 = yindex % 3
y1 = yindex // 3
tmp0 = tl.load(in_ptr0 + (x2 + 4096 * y3), ymask, eviction_policy=
'evict_last')
tl.store(out_ptr0 + (y0 + 3 * x2 + 12288 * y1), tmp0, ymask)
@triton.jit
def triton_poi_fused_convolution_1(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl
.constexpr):
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
tl.full([XBLOCK], True, tl.int1)
x2 = xindex
x0 = xindex % 100
tmp0 = tl.load(in_out_ptr0 + x2, None)
tmp1 = tl.load(in_ptr0 + x0, None, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + x2, tmp2, None)
@triton.jit
def triton_per_fused_native_group_norm_2(in_ptr0, out_ptr0, out_ptr1,
out_ptr2, xnumel, rnumel, XBLOCK: tl.constexpr):
xnumel = 12800
RBLOCK: tl.constexpr = 128
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
r3 = rindex
x0 = xindex % 50
x1 = xindex // 50 % 64
x2 = xindex // 3200
x4 = xindex
tmp0 = tl.load(in_ptr0 + (2 * x0 + 100 * ((r3 + 128 * x1) % 4096) +
409600 * x2 + (r3 + 128 * x1) // 4096), xmask, eviction_policy=
'evict_last', other=0.0)
tmp1 = tl.broadcast_to(tmp0, [XBLOCK, RBLOCK])
tl.where(xmask, tmp1, 0)
tmp4 = tl.broadcast_to(tmp1, [XBLOCK, RBLOCK])
tmp6 = tl.where(xmask, tmp4, 0)
tmp7 = tl.sum(tmp6, 1)[:, None]
tmp8 = tl.full([XBLOCK, 1], 128, tl.int32)
tmp9 = tmp8.to(tl.float32)
tmp10 = tmp7 / tmp9
tmp11 = tmp1 - tmp10
tmp12 = tmp11 * tmp11
tmp13 = tl.broadcast_to(tmp12, [XBLOCK, RBLOCK])
tmp15 = tl.where(xmask, tmp13, 0)
tmp16 = tl.sum(tmp15, 1)[:, None]
tl.store(out_ptr0 + x4, tmp10, xmask)
tl.store(out_ptr1 + x4, tmp16, xmask)
tl.store(out_ptr2 + x4, tmp9, xmask)
@triton.jit
def triton_per_fused_native_group_norm_3(in_ptr0, in_ptr1, in_ptr2,
out_ptr0, out_ptr1, out_ptr2, xnumel, rnumel, XBLOCK: tl.constexpr):
xnumel = 200
RBLOCK: tl.constexpr = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
r2 = rindex
x0 = xindex % 50
x1 = xindex // 50
x3 = xindex
tmp0 = tl.load(in_ptr0 + (x0 + 50 * r2 + 3200 * x1), xmask, other=0.0)
tmp1 = tl.load(in_ptr1 + (x0 + 50 * r2 + 3200 * x1), xmask, other=0.0)
tmp2 = tl.load(in_ptr2 + (x0 + 50 * r2 + 3200 * x1), xmask, other=0.0)
tmp3 = tl.broadcast_to(tmp0, [XBLOCK, RBLOCK])
tmp4 = tl.broadcast_to(tmp1, [XBLOCK, RBLOCK])
tmp5 = tl.broadcast_to(tmp2, [XBLOCK, RBLOCK])
tmp7 = tl.where(xmask, tmp3, 0)
tmp8 = tl.where(xmask, tmp4, 0)
tmp9 = tl.where(xmask, tmp5, 0)
tmp10, tmp11, tmp12 = triton_helpers.welford(tmp7, tmp8, tmp9, 1)
tmp13 = tmp10[:, None]
tmp14 = tmp11[:, None]
tmp15 = tmp12[:, None]
tl.store(out_ptr0 + x3, tmp13, xmask)
tl.store(out_ptr1 + x3, tmp14, xmask)
tl.store(out_ptr2 + x3, tmp15, xmask)
@triton.jit
def triton_per_fused_native_group_norm_4(in_ptr0, in_ptr1, in_ptr2,
out_ptr0, out_ptr1, out_ptr2, xnumel, rnumel, XBLOCK: tl.constexpr):
xnumel = 100
RBLOCK: tl.constexpr = 2
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
r1 = rindex
x0 = xindex
tmp0 = tl.load(in_ptr0 + (r1 + 2 * x0), xmask, other=0.0)
tmp1 = tl.load(in_ptr1 + (r1 + 2 * x0), xmask, other=0.0)
tmp2 = tl.load(in_ptr2 + (r1 + 2 * x0), xmask, other=0.0)
tmp3 = tl.broadcast_to(tmp0, [XBLOCK, RBLOCK])
tmp4 = tl.broadcast_to(tmp1, [XBLOCK, RBLOCK])
tmp5 = tl.broadcast_to(tmp2, [XBLOCK, RBLOCK])
tmp7 = tl.where(xmask, tmp3, 0)
tmp8 = tl.where(xmask, tmp4, 0)
tmp9 = tl.where(xmask, tmp5, 0)
tmp10, tmp11, tmp12 = triton_helpers.welford(tmp7, tmp8, tmp9, 1)
tmp13 = tmp10[:, None]
tmp14 = tmp11[:, None]
tmp12[:, None]
tmp16 = 16384.0
tmp17 = tmp14 / tmp16
tmp18 = 1e-05
tmp19 = tmp17 + tmp18
tmp20 = libdevice.rsqrt(tmp19)
tl.store(out_ptr2 + x0, tmp20, xmask)
tl.store(out_ptr0 + x0, tmp13, xmask)
tl.store(out_ptr1 + x0, tmp14, xmask)
@triton.jit
def triton_poi_fused_native_group_norm_relu_threshold_backward_5(in_ptr0,
in_ptr1, in_ptr2, in_ptr3, in_ptr4, out_ptr0, out_ptr1, ynumel, xnumel,
YBLOCK: tl.constexpr, XBLOCK: tl.constexpr):
ynumel = 400
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
tl.full([XBLOCK, YBLOCK], True, tl.int1)
x2 = xindex
y0 = yindex % 100
y1 = yindex // 100
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + 100 * x2 + 409600 * y1), ymask,
eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + y3 // 4, ymask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr2 + y3 // 4, ymask, eviction_policy='evict_last')
tmp10 = tl.load(in_ptr3 + y0, ymask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr4 + y0, ymask, eviction_policy='evict_last')
tmp2 = tmp0 - tmp1
tmp4 = 16384.0
tmp5 = tmp3 / tmp4
tmp6 = 1e-05
tmp7 = tmp5 + tmp6
tmp8 = libdevice.rsqrt(tmp7)
tmp9 = tmp2 * tmp8
tmp11 = tmp9 * tmp10
tmp13 = tmp11 + tmp12
tmp14 = tl.full([1, 1], 0, tl.int32)
tmp15 = triton_helpers.maximum(tmp14, tmp13)
tmp16 = 0.0
tmp17 = tmp15 <= tmp16
tl.store(out_ptr0 + (x2 + 4096 * y3), tmp15, ymask)
tl.store(out_ptr1 + (y0 + 100 * x2 + 409600 * y1), tmp17, ymask)
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5 = args
args.clear()
assert_size_stride(primals_1, (100, 3, 1, 1), (3, 1, 1, 1))
assert_size_stride(primals_2, (100,), (1,))
assert_size_stride(primals_3, (4, 3, 64, 64), (12288, 4096, 64, 1))
assert_size_stride(primals_4, (100,), (1,))
assert_size_stride(primals_5, (100,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 3, 64, 64), (12288, 1, 192, 3), torch
.float32)
get_raw_stream(0)
triton_poi_fused_0[grid(12, 4096)](primals_3, buf0, 12, 4096,
XBLOCK=64, YBLOCK=16, num_warps=4, num_stages=1)
del primals_3
buf1 = extern_kernels.convolution(buf0, primals_1, stride=(1, 1),
padding=(0, 0), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf1, (4, 100, 64, 64), (409600, 1, 6400, 100))
buf2 = buf1
del buf1
triton_poi_fused_convolution_1[grid(1638400)](buf2, primals_2,
1638400, XBLOCK=512, num_warps=8, num_stages=1)
del primals_2
buf3 = empty_strided_cuda((4, 25, 1, 1, 2, 64), (3200, 2, 12800,
12800, 1, 50), torch.float32)
buf4 = empty_strided_cuda((4, 25, 1, 1, 2, 64), (3200, 2, 12800,
12800, 1, 50), torch.float32)
buf5 = empty_strided_cuda((4, 25, 1, 1, 2, 64), (3200, 2, 12800,
12800, 1, 50), torch.float32)
triton_per_fused_native_group_norm_2[grid(12800)](buf2, buf3, buf4,
buf5, 12800, 128, XBLOCK=8, num_warps=8, num_stages=1)
buf6 = empty_strided_cuda((4, 25, 1, 1, 2), (50, 2, 200, 200, 1),
torch.float32)
buf7 = empty_strided_cuda((4, 25, 1, 1, 2), (50, 2, 200, 200, 1),
torch.float32)
buf8 = empty_strided_cuda((4, 25, 1, 1, 2), (50, 2, 200, 200, 1),
torch.float32)
triton_per_fused_native_group_norm_3[grid(200)](buf3, buf4, buf5,
buf6, buf7, buf8, 200, 64, XBLOCK=8, num_warps=4, num_stages=1)
del buf3
del buf4
del buf5
buf9 = empty_strided_cuda((4, 25, 1, 1), (25, 1, 100, 100), torch.
float32)
buf10 = empty_strided_cuda((4, 25, 1, 1), (25, 1, 100, 100), torch.
float32)
buf12 = empty_strided_cuda((4, 25, 1, 1), (25, 1, 100, 100), torch.
float32)
triton_per_fused_native_group_norm_4[grid(100)](buf6, buf7, buf8,
buf9, buf10, buf12, 100, 2, XBLOCK=8, num_warps=2, num_stages=1)
del buf6
del buf7
del buf8
buf13 = empty_strided_cuda((4, 100, 64, 64), (409600, 4096, 64, 1),
torch.float32)
buf14 = empty_strided_cuda((4, 100, 64, 64), (409600, 1, 6400, 100),
torch.bool)
triton_poi_fused_native_group_norm_relu_threshold_backward_5[grid(
400, 4096)](buf2, buf9, buf10, primals_4, primals_5, buf13,
buf14, 400, 4096, XBLOCK=32, YBLOCK=32, num_warps=4, num_stages=1)
del buf10
del primals_5
return buf13, primals_1, buf0, primals_4, buf2, reinterpret_tensor(buf9,
(4, 25), (25, 1), 0), reinterpret_tensor(buf12, (4, 25), (25, 1), 0
), buf14
class DynamicPreHeadNew(nn.Module):
def __init__(self, in_dim=3, embed_dim=100, kernel_size=1):
super(DynamicPreHeadNew, self).__init__()
self.conv = nn.Conv2d(in_dim, embed_dim, kernel_size=kernel_size,
stride=1, padding=int((kernel_size - 1) / 2))
self.bn = nn.GroupNorm(int(embed_dim / 4), embed_dim)
self.relu = nn.ReLU(True)
nn.init.kaiming_normal_(self.conv.weight, mode='fan_out',
nonlinearity='relu')
def forward(self, input_0):
primals_1 = self.conv.weight
primals_2 = self.conv.bias
primals_4 = self.bn.weight
primals_5 = self.bn.bias
primals_3 = input_0
output = call([primals_1, primals_2, primals_3, primals_4, primals_5])
return output[0]
| huanglf714/COMatchNet | DynamicPreHead | false | 6,838 | [
"Apache-2.0"
] | 1 | 79023f5be65d354eb9bdac026d7e0d73110bc4aa | https://github.com/huanglf714/COMatchNet/tree/79023f5be65d354eb9bdac026d7e0d73110bc4aa | import torch
import torch.nn as nn
class Model(nn.Module):
def __init__(self, in_dim=3, embed_dim=100, kernel_size=1):
super().__init__()
self.conv = nn.Conv2d(in_dim, embed_dim, kernel_size=kernel_size,
stride=1, padding=int((kernel_size - 1) / 2))
self.bn = nn.GroupNorm(int(embed_dim / 4), embed_dim)
self.relu = nn.ReLU(True)
nn.init.kaiming_normal_(self.conv.weight, mode='fan_out',
nonlinearity='relu')
def forward(self, x):
x = self.conv(x)
x = self.bn(x)
x = self.relu(x)
return x
def get_inputs():
return [torch.rand([4, 3, 64, 64])]
def get_init_inputs():
return []
|
NormalizationLayer | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/5d/c5dhk7bkuqo4u6hoo6g4wm36lekpofplwrblvshjxay6uhmurm2w.py
# Topologically Sorted Source Nodes: [mul, features], Original ATen: [aten.mul, aten.div]
# Source node to ATen node mapping:
# features => div
# mul => mul
# Graph fragment:
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%primals_1, %primals_2), kwargs = {})
# %div : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%mul, %expand), kwargs = {})
triton_poi_fused_div_mul_0 = async_compile.triton('triton_poi_fused_div_mul_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_div_mul_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 6, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_div_mul_0(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x0 = xindex % 16
x2 = (xindex // 64)
tmp0 = tl.load(in_ptr0 + (0))
tmp1 = tl.broadcast_to(tmp0, [XBLOCK])
tmp2 = tl.load(in_ptr1 + (x3), xmask)
tmp4 = tl.load(in_ptr1 + (x0 + (64*x2)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr1 + (16 + x0 + (64*x2)), xmask, eviction_policy='evict_last')
tmp9 = tl.load(in_ptr1 + (32 + x0 + (64*x2)), xmask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr1 + (48 + x0 + (64*x2)), xmask, eviction_policy='evict_last')
tmp3 = tmp1 * tmp2
tmp5 = tmp4 * tmp4
tmp7 = tmp6 * tmp6
tmp8 = tmp5 + tmp7
tmp10 = tmp9 * tmp9
tmp11 = tmp8 + tmp10
tmp13 = tmp12 * tmp12
tmp14 = tmp11 + tmp13
tmp15 = libdevice.sqrt(tmp14)
tmp16 = tmp3 / tmp15
tl.store(out_ptr0 + (x3), tmp16, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2 = args
args.clear()
assert_size_stride(primals_1, (1, ), (1, ))
assert_size_stride(primals_2, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [mul, features], Original ATen: [aten.mul, aten.div]
stream0 = get_raw_stream(0)
triton_poi_fused_div_mul_0.run(primals_1, primals_2, buf0, 256, grid=grid(256), stream=stream0)
del primals_1
return (buf0, primals_2, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((1, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.utils.data
class NormalizationLayer(torch.nn.Module):
"""Class for normalization layer."""
def __init__(self, normalize_scale=1.0, learn_scale=True):
super(NormalizationLayer, self).__init__()
self.norm_s = float(normalize_scale)
if learn_scale:
self.norm_s = torch.nn.Parameter(torch.FloatTensor((self.norm_s,)))
def forward(self, x):
features = self.norm_s * x / torch.norm(x, dim=1, keepdim=True
).expand_as(x)
return features
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {}]
| import torch
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime.triton_helpers import libdevice
import torch.utils.data
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_poi_fused_div_mul_0(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK:
tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x0 = xindex % 16
x2 = xindex // 64
tmp0 = tl.load(in_ptr0 + 0)
tmp1 = tl.broadcast_to(tmp0, [XBLOCK])
tmp2 = tl.load(in_ptr1 + x3, xmask)
tmp4 = tl.load(in_ptr1 + (x0 + 64 * x2), xmask, eviction_policy=
'evict_last')
tmp6 = tl.load(in_ptr1 + (16 + x0 + 64 * x2), xmask, eviction_policy=
'evict_last')
tmp9 = tl.load(in_ptr1 + (32 + x0 + 64 * x2), xmask, eviction_policy=
'evict_last')
tmp12 = tl.load(in_ptr1 + (48 + x0 + 64 * x2), xmask, eviction_policy=
'evict_last')
tmp3 = tmp1 * tmp2
tmp5 = tmp4 * tmp4
tmp7 = tmp6 * tmp6
tmp8 = tmp5 + tmp7
tmp10 = tmp9 * tmp9
tmp11 = tmp8 + tmp10
tmp13 = tmp12 * tmp12
tmp14 = tmp11 + tmp13
tmp15 = libdevice.sqrt(tmp14)
tmp16 = tmp3 / tmp15
tl.store(out_ptr0 + x3, tmp16, xmask)
def call(args):
primals_1, primals_2 = args
args.clear()
assert_size_stride(primals_1, (1,), (1,))
assert_size_stride(primals_2, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_div_mul_0[grid(256)](primals_1, primals_2, buf0,
256, XBLOCK=128, num_warps=4, num_stages=1)
del primals_1
return buf0, primals_2
class NormalizationLayerNew(torch.nn.Module):
"""Class for normalization layer."""
def __init__(self, normalize_scale=1.0, learn_scale=True):
super(NormalizationLayerNew, self).__init__()
self.norm_s = float(normalize_scale)
if learn_scale:
self.norm_s = torch.nn.Parameter(torch.FloatTensor((self.norm_s,)))
def forward(self, input_0):
primals_1 = self.norm_s
primals_2 = input_0
output = call([primals_1, primals_2])
return output[0]
| huynhtruc0309/tirg | NormalizationLayer | false | 6,839 | [
"Apache-2.0"
] | 1 | 14ac6dcb41624729a6f4144a7c9e7899074f0eec | https://github.com/huynhtruc0309/tirg/tree/14ac6dcb41624729a6f4144a7c9e7899074f0eec | import torch
import torch.utils.data
class Model(torch.nn.Module):
"""Class for normalization layer."""
def __init__(self, normalize_scale=1.0, learn_scale=True):
super().__init__()
self.norm_s = float(normalize_scale)
if learn_scale:
self.norm_s = torch.nn.Parameter(torch.FloatTensor((self.norm_s,)))
def forward(self, x):
features = self.norm_s * x / torch.norm(x, dim=1, keepdim=True
).expand_as(x)
return features
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return []
|
decoder2 | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/mb/cmb72vxh36b4k6lvmt4562lj3nrqtpyzst2qbon2yqx22gdjfa7x.py
# Topologically Sorted Source Nodes: [z], Original ATen: [aten.sigmoid]
# Source node to ATen node mapping:
# z => sigmoid
# Graph fragment:
# %sigmoid : [num_users=1] = call_function[target=torch.ops.aten.sigmoid.default](args = (%mm,), kwargs = {})
triton_poi_fused_sigmoid_0 = async_compile.triton('triton_poi_fused_sigmoid_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_sigmoid_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_sigmoid_0(in_out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_out_ptr0 + (x0), xmask)
tmp1 = tl.sigmoid(tmp0)
tl.store(in_out_ptr0 + (x0), tmp1, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, arg1_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4), (4, 1))
assert_size_stride(arg1_1, (4, 4), (4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [mm], Original ATen: [aten.mm]
extern_kernels.mm(arg0_1, reinterpret_tensor(arg1_1, (4, 4), (1, 4), 0), out=buf0)
del arg0_1
del arg1_1
buf1 = buf0; del buf0 # reuse
# Topologically Sorted Source Nodes: [z], Original ATen: [aten.sigmoid]
stream0 = get_raw_stream(0)
triton_poi_fused_sigmoid_0.run(buf1, 16, grid=grid(16), stream=stream0)
return (buf1, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
arg1_1 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([arg0_1, arg1_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
class decoder2(nn.Module):
def __init__(self, dropout=0.5, act=torch.sigmoid):
super(decoder2, self).__init__()
self.dropout = nn.Dropout(dropout)
self.act = act
def forward(self, z_node, z_hyperedge):
z_node_ = self.dropout(z_node)
z_hyperedge_ = self.dropout(z_hyperedge)
z = self.act(z_node_.mm(z_hyperedge_.t()))
return z
def get_inputs():
return [torch.rand([4, 4]), torch.rand([4, 4])]
def get_init_inputs():
return [[], {}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_sigmoid_0(in_out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_out_ptr0 + x0, xmask)
tmp1 = tl.sigmoid(tmp0)
tl.store(in_out_ptr0 + x0, tmp1, xmask)
def call(args):
arg0_1, arg1_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4), (4, 1))
assert_size_stride(arg1_1, (4, 4), (4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.mm(arg0_1, reinterpret_tensor(arg1_1, (4, 4), (1, 4),
0), out=buf0)
del arg0_1
del arg1_1
buf1 = buf0
del buf0
get_raw_stream(0)
triton_poi_fused_sigmoid_0[grid(16)](buf1, 16, XBLOCK=16, num_warps
=1, num_stages=1)
return buf1,
class decoder2New(nn.Module):
def __init__(self, dropout=0.5, act=torch.sigmoid):
super(decoder2New, self).__init__()
self.dropout = nn.Dropout(dropout)
self.act = act
def forward(self, input_0, input_1):
arg0_1 = input_0
arg1_1 = input_1
output = call([arg0_1, arg1_1])
return output[0]
| iMoonLab/HHDTI | decoder2 | false | 6,840 | [
"MIT"
] | 1 | b2dd0e78818888e676afc91af1425dada5b3258a | https://github.com/iMoonLab/HHDTI/tree/b2dd0e78818888e676afc91af1425dada5b3258a | import torch
import torch.nn as nn
class Model(nn.Module):
def __init__(self, dropout=0.5, act=torch.sigmoid):
super().__init__()
self.dropout = nn.Dropout(dropout)
self.act = act
def forward(self, z_node, z_hyperedge):
z_node_ = self.dropout(z_node)
z_hyperedge_ = self.dropout(z_hyperedge)
z = self.act(z_node_.mm(z_hyperedge_.t()))
return z
def get_inputs():
return [torch.rand([4, 4]), torch.rand([4, 4])]
def get_init_inputs():
return []
|
node_encoder | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/iy/ciydlc67cqubwi3mjoecv367dejmlq7ltvvfugle7nzw4qyevqog.py
# Topologically Sorted Source Nodes: [add, z1], Original ATen: [aten.add, aten.tanh]
# Source node to ATen node mapping:
# add => add
# z1 => tanh
# Graph fragment:
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mm, %primals_3), kwargs = {})
# %tanh : [num_users=1] = call_function[target=torch.ops.aten.tanh.default](args = (%add,), kwargs = {})
triton_poi_fused_add_tanh_0 = async_compile.triton('triton_poi_fused_add_tanh_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_tanh_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_tanh_0(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = libdevice.tanh(tmp2)
tl.store(in_out_ptr0 + (x2), tmp3, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (4, 4), (4, 1))
assert_size_stride(primals_2, (4, 4), (4, 1))
assert_size_stride(primals_3, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [mm], Original ATen: [aten.mm]
extern_kernels.mm(primals_1, primals_2, out=buf0)
del primals_2
buf1 = buf0; del buf0 # reuse
# Topologically Sorted Source Nodes: [add, z1], Original ATen: [aten.add, aten.tanh]
stream0 = get_raw_stream(0)
triton_poi_fused_add_tanh_0.run(buf1, primals_3, 16, grid=grid(16), stream=stream0)
del primals_3
return (buf1, buf1, reinterpret_tensor(primals_1, (4, 4), (1, 4), 0), )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.nn.functional as F
class node_encoder(nn.Module):
def __init__(self, num_in_node, num_hidden, dropout, act=F.tanh):
super(node_encoder, self).__init__()
self.num_in_node = num_in_node
self.num_hidden = num_hidden
self.dropout = dropout
self.act = act
self.W1 = nn.Parameter(torch.zeros(size=(self.num_in_node, self.
num_hidden), dtype=torch.float))
nn.init.xavier_uniform_(self.W1.data, gain=1.414)
self.b1 = nn.Parameter(torch.zeros(self.num_hidden, dtype=torch.float))
def forward(self, H):
z1 = self.act(H.mm(self.W1) + self.b1)
return z1
def __repr__(self):
return self.__class__.__name__ + '(' + str(self.num_in_node
) + ' -> ' + str(self.num_hidden)
def get_inputs():
return [torch.rand([4, 4])]
def get_init_inputs():
return [[], {'num_in_node': 4, 'num_hidden': 4, 'dropout': 0.5}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime.triton_helpers import libdevice
import torch.nn as nn
import torch.nn.functional as F
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_add_tanh_0(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl.
constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = libdevice.tanh(tmp2)
tl.store(in_out_ptr0 + x2, tmp3, xmask)
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (4, 4), (4, 1))
assert_size_stride(primals_2, (4, 4), (4, 1))
assert_size_stride(primals_3, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.mm(primals_1, primals_2, out=buf0)
del primals_2
buf1 = buf0
del buf0
get_raw_stream(0)
triton_poi_fused_add_tanh_0[grid(16)](buf1, primals_3, 16, XBLOCK=
16, num_warps=1, num_stages=1)
del primals_3
return buf1, buf1, reinterpret_tensor(primals_1, (4, 4), (1, 4), 0)
class node_encoderNew(nn.Module):
def __init__(self, num_in_node, num_hidden, dropout, act=F.tanh):
super(node_encoderNew, self).__init__()
self.num_in_node = num_in_node
self.num_hidden = num_hidden
self.dropout = dropout
self.act = act
self.W1 = nn.Parameter(torch.zeros(size=(self.num_in_node, self.
num_hidden), dtype=torch.float))
nn.init.xavier_uniform_(self.W1.data, gain=1.414)
self.b1 = nn.Parameter(torch.zeros(self.num_hidden, dtype=torch.float))
def __repr__(self):
return self.__class__.__name__ + '(' + str(self.num_in_node
) + ' -> ' + str(self.num_hidden)
def forward(self, input_0):
primals_1 = self.W1
primals_3 = self.b1
primals_2 = input_0
output = call([primals_1, primals_2, primals_3])
return output[0]
| iMoonLab/HHDTI | node_encoder | false | 6,841 | [
"MIT"
] | 1 | b2dd0e78818888e676afc91af1425dada5b3258a | https://github.com/iMoonLab/HHDTI/tree/b2dd0e78818888e676afc91af1425dada5b3258a | import torch
import torch.nn as nn
import torch.nn.functional as F
class Model(nn.Module):
def __init__(self, num_in_node, num_hidden, dropout, act=F.tanh):
super().__init__()
self.num_in_node = num_in_node
self.num_hidden = num_hidden
self.dropout = dropout
self.act = act
self.W1 = nn.Parameter(torch.zeros(size=(self.num_in_node, self.
num_hidden), dtype=torch.float))
nn.init.xavier_uniform_(self.W1.data, gain=1.414)
self.b1 = nn.Parameter(torch.zeros(self.num_hidden, dtype=torch.float))
def forward(self, H):
z1 = self.act(H.mm(self.W1) + self.b1)
return z1
def __repr__(self):
return self.__class__.__name__ + '(' + str(self.num_in_node
) + ' -> ' + str(self.num_hidden)
def get_inputs():
return [torch.rand([4, 4])]
def get_init_inputs():
return [4, 4, 0.5]
|
hyperedge_encoder | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/iy/ciydlc67cqubwi3mjoecv367dejmlq7ltvvfugle7nzw4qyevqog.py
# Topologically Sorted Source Nodes: [add, z1], Original ATen: [aten.add, aten.tanh]
# Source node to ATen node mapping:
# add => add
# z1 => tanh
# Graph fragment:
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mm, %primals_3), kwargs = {})
# %tanh : [num_users=1] = call_function[target=torch.ops.aten.tanh.default](args = (%add,), kwargs = {})
triton_poi_fused_add_tanh_0 = async_compile.triton('triton_poi_fused_add_tanh_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_tanh_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_tanh_0(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = libdevice.tanh(tmp2)
tl.store(in_out_ptr0 + (x2), tmp3, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (4, 4), (4, 1))
assert_size_stride(primals_2, (4, 4), (4, 1))
assert_size_stride(primals_3, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [mm], Original ATen: [aten.mm]
extern_kernels.mm(primals_2, primals_1, out=buf0)
del primals_1
buf1 = buf0; del buf0 # reuse
# Topologically Sorted Source Nodes: [add, z1], Original ATen: [aten.add, aten.tanh]
stream0 = get_raw_stream(0)
triton_poi_fused_add_tanh_0.run(buf1, primals_3, 16, grid=grid(16), stream=stream0)
del primals_3
return (buf1, buf1, reinterpret_tensor(primals_2, (4, 4), (1, 4), 0), )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.nn.functional as F
class hyperedge_encoder(nn.Module):
def __init__(self, num_in_edge, num_hidden, dropout, act=F.tanh):
super(hyperedge_encoder, self).__init__()
self.num_in_edge = num_in_edge
self.num_hidden = num_hidden
self.dropout = dropout
self.act = act
self.W1 = nn.Parameter(torch.zeros(size=(self.num_in_edge, self.
num_hidden), dtype=torch.float))
nn.init.xavier_uniform_(self.W1.data, gain=1.414)
self.b1 = nn.Parameter(torch.zeros(self.num_hidden, dtype=torch.float))
def forward(self, H_T):
z1 = self.act(torch.mm(H_T, self.W1) + self.b1)
return z1
def __repr__(self):
return self.__class__.__name__ + '(' + str(self.num_in_edge
) + ' -> ' + str(self.num_hidden)
def get_inputs():
return [torch.rand([4, 4])]
def get_init_inputs():
return [[], {'num_in_edge': 4, 'num_hidden': 4, 'dropout': 0.5}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime.triton_helpers import libdevice
import torch.nn as nn
import torch.nn.functional as F
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_add_tanh_0(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl.
constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = libdevice.tanh(tmp2)
tl.store(in_out_ptr0 + x2, tmp3, xmask)
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (4, 4), (4, 1))
assert_size_stride(primals_2, (4, 4), (4, 1))
assert_size_stride(primals_3, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.mm(primals_2, primals_1, out=buf0)
del primals_1
buf1 = buf0
del buf0
get_raw_stream(0)
triton_poi_fused_add_tanh_0[grid(16)](buf1, primals_3, 16, XBLOCK=
16, num_warps=1, num_stages=1)
del primals_3
return buf1, buf1, reinterpret_tensor(primals_2, (4, 4), (1, 4), 0)
class hyperedge_encoderNew(nn.Module):
def __init__(self, num_in_edge, num_hidden, dropout, act=F.tanh):
super(hyperedge_encoderNew, self).__init__()
self.num_in_edge = num_in_edge
self.num_hidden = num_hidden
self.dropout = dropout
self.act = act
self.W1 = nn.Parameter(torch.zeros(size=(self.num_in_edge, self.
num_hidden), dtype=torch.float))
nn.init.xavier_uniform_(self.W1.data, gain=1.414)
self.b1 = nn.Parameter(torch.zeros(self.num_hidden, dtype=torch.float))
def __repr__(self):
return self.__class__.__name__ + '(' + str(self.num_in_edge
) + ' -> ' + str(self.num_hidden)
def forward(self, input_0):
primals_1 = self.W1
primals_3 = self.b1
primals_2 = input_0
output = call([primals_1, primals_2, primals_3])
return output[0]
| iMoonLab/HHDTI | hyperedge_encoder | false | 6,842 | [
"MIT"
] | 1 | b2dd0e78818888e676afc91af1425dada5b3258a | https://github.com/iMoonLab/HHDTI/tree/b2dd0e78818888e676afc91af1425dada5b3258a | import torch
import torch.nn as nn
import torch.nn.functional as F
class Model(nn.Module):
def __init__(self, num_in_edge, num_hidden, dropout, act=F.tanh):
super().__init__()
self.num_in_edge = num_in_edge
self.num_hidden = num_hidden
self.dropout = dropout
self.act = act
self.W1 = nn.Parameter(torch.zeros(size=(self.num_in_edge, self.
num_hidden), dtype=torch.float))
nn.init.xavier_uniform_(self.W1.data, gain=1.414)
self.b1 = nn.Parameter(torch.zeros(self.num_hidden, dtype=torch.float))
def forward(self, H_T):
z1 = self.act(torch.mm(H_T, self.W1) + self.b1)
return z1
def __repr__(self):
return self.__class__.__name__ + '(' + str(self.num_in_edge
) + ' -> ' + str(self.num_hidden)
def get_inputs():
return [torch.rand([4, 4])]
def get_init_inputs():
return [4, 4, 0.5]
|
kl_loss | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/4d/c4dd33o7csaipr5j7mpc75mdyfwkdwifxk5nhggvn2k3fzu7jxgp.py
# Topologically Sorted Source Nodes: [mul, add, pow_1, sub, exp, pow_2, sub_1, sum_1, mean, kl_node, mul_2, add_1, pow_3, sub_2, exp_1, pow_4, sub_3, sum_2, mean_1, kl_edge, kl], Original ATen: [aten.mul, aten.add, aten.pow, aten.sub, aten.exp, aten.sum, aten.mean]
# Source node to ATen node mapping:
# add => add
# add_1 => add_1
# exp => exp
# exp_1 => exp_1
# kl => add_2
# kl_edge => mul_3
# kl_node => mul_1
# mean => mean
# mean_1 => mean_1
# mul => mul
# mul_2 => mul_2
# pow_1 => pow_1
# pow_2 => pow_2
# pow_3 => pow_3
# pow_4 => pow_4
# sub => sub
# sub_1 => sub_1
# sub_2 => sub_2
# sub_3 => sub_3
# sum_1 => sum_1
# sum_2 => sum_2
# Graph fragment:
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%arg0_1, 2), kwargs = {})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul, 1), kwargs = {})
# %pow_1 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%arg1_1, 2), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%add, %pow_1), kwargs = {})
# %exp : [num_users=1] = call_function[target=torch.ops.aten.exp.default](args = (%arg0_1,), kwargs = {})
# %pow_2 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%exp, 2), kwargs = {})
# %sub_1 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%sub, %pow_2), kwargs = {})
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%sub_1, [1]), kwargs = {})
# %mean : [num_users=1] = call_function[target=torch.ops.aten.mean.default](args = (%sum_1,), kwargs = {})
# %mul_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%mean, -0.125), kwargs = {})
# %mul_2 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%arg2_1, 2), kwargs = {})
# %add_1 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul_2, 1), kwargs = {})
# %pow_3 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%arg3_1, 2), kwargs = {})
# %sub_2 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%add_1, %pow_3), kwargs = {})
# %exp_1 : [num_users=1] = call_function[target=torch.ops.aten.exp.default](args = (%arg2_1,), kwargs = {})
# %pow_4 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%exp_1, 2), kwargs = {})
# %sub_3 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%sub_2, %pow_4), kwargs = {})
# %sum_2 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%sub_3, [1]), kwargs = {})
# %mean_1 : [num_users=1] = call_function[target=torch.ops.aten.mean.default](args = (%sum_2,), kwargs = {})
# %mul_3 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%mean_1, -0.125), kwargs = {})
# %add_2 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul_1, %mul_3), kwargs = {})
triton_per_fused_add_exp_mean_mul_pow_sub_sum_0 = async_compile.triton('triton_per_fused_add_exp_mean_mul_pow_sub_sum_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[1, 64],
reduction_hint=ReductionHint.INNER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: '*fp32', 5: 'i32', 6: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {5: 1}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 6), equal_to_1=(5,))]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused_add_exp_mean_mul_pow_sub_sum_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 16, 'num_reduction': 2, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused_add_exp_mean_mul_pow_sub_sum_0(in_out_ptr0, in_ptr0, in_ptr1, in_ptr2, in_ptr3, xnumel, rnumel, XBLOCK : tl.constexpr):
xnumel = 1
rnumel = 64
RBLOCK: tl.constexpr = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[None, :]
roffset = 0
rmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
r0 = rindex % 16
r1 = (rindex // 16)
r2 = rindex
tmp0 = tl.load(in_ptr0 + (r0 + (64*r1)), None)
tmp5 = tl.load(in_ptr1 + (r0 + (64*r1)), None)
tmp11 = tl.load(in_ptr0 + (16 + r0 + (64*r1)), None)
tmp14 = tl.load(in_ptr1 + (16 + r0 + (64*r1)), None)
tmp21 = tl.load(in_ptr0 + (32 + r0 + (64*r1)), None)
tmp24 = tl.load(in_ptr1 + (32 + r0 + (64*r1)), None)
tmp31 = tl.load(in_ptr0 + (48 + r0 + (64*r1)), None)
tmp34 = tl.load(in_ptr1 + (48 + r0 + (64*r1)), None)
tmp44 = tl.load(in_ptr2 + (r0 + (64*r1)), None)
tmp47 = tl.load(in_ptr3 + (r0 + (64*r1)), None)
tmp53 = tl.load(in_ptr2 + (16 + r0 + (64*r1)), None)
tmp56 = tl.load(in_ptr3 + (16 + r0 + (64*r1)), None)
tmp63 = tl.load(in_ptr2 + (32 + r0 + (64*r1)), None)
tmp66 = tl.load(in_ptr3 + (32 + r0 + (64*r1)), None)
tmp73 = tl.load(in_ptr2 + (48 + r0 + (64*r1)), None)
tmp76 = tl.load(in_ptr3 + (48 + r0 + (64*r1)), None)
tmp1 = 2.0
tmp2 = tmp0 * tmp1
tmp3 = 1.0
tmp4 = tmp2 + tmp3
tmp6 = tmp5 * tmp5
tmp7 = tmp4 - tmp6
tmp8 = tl_math.exp(tmp0)
tmp9 = tmp8 * tmp8
tmp10 = tmp7 - tmp9
tmp12 = tmp11 * tmp1
tmp13 = tmp12 + tmp3
tmp15 = tmp14 * tmp14
tmp16 = tmp13 - tmp15
tmp17 = tl_math.exp(tmp11)
tmp18 = tmp17 * tmp17
tmp19 = tmp16 - tmp18
tmp20 = tmp10 + tmp19
tmp22 = tmp21 * tmp1
tmp23 = tmp22 + tmp3
tmp25 = tmp24 * tmp24
tmp26 = tmp23 - tmp25
tmp27 = tl_math.exp(tmp21)
tmp28 = tmp27 * tmp27
tmp29 = tmp26 - tmp28
tmp30 = tmp20 + tmp29
tmp32 = tmp31 * tmp1
tmp33 = tmp32 + tmp3
tmp35 = tmp34 * tmp34
tmp36 = tmp33 - tmp35
tmp37 = tl_math.exp(tmp31)
tmp38 = tmp37 * tmp37
tmp39 = tmp36 - tmp38
tmp40 = tmp30 + tmp39
tmp41 = tl.broadcast_to(tmp40, [XBLOCK, RBLOCK])
tmp43 = tl.sum(tmp41, 1)[:, None]
tmp45 = tmp44 * tmp1
tmp46 = tmp45 + tmp3
tmp48 = tmp47 * tmp47
tmp49 = tmp46 - tmp48
tmp50 = tl_math.exp(tmp44)
tmp51 = tmp50 * tmp50
tmp52 = tmp49 - tmp51
tmp54 = tmp53 * tmp1
tmp55 = tmp54 + tmp3
tmp57 = tmp56 * tmp56
tmp58 = tmp55 - tmp57
tmp59 = tl_math.exp(tmp53)
tmp60 = tmp59 * tmp59
tmp61 = tmp58 - tmp60
tmp62 = tmp52 + tmp61
tmp64 = tmp63 * tmp1
tmp65 = tmp64 + tmp3
tmp67 = tmp66 * tmp66
tmp68 = tmp65 - tmp67
tmp69 = tl_math.exp(tmp63)
tmp70 = tmp69 * tmp69
tmp71 = tmp68 - tmp70
tmp72 = tmp62 + tmp71
tmp74 = tmp73 * tmp1
tmp75 = tmp74 + tmp3
tmp77 = tmp76 * tmp76
tmp78 = tmp75 - tmp77
tmp79 = tl_math.exp(tmp73)
tmp80 = tmp79 * tmp79
tmp81 = tmp78 - tmp80
tmp82 = tmp72 + tmp81
tmp83 = tl.broadcast_to(tmp82, [XBLOCK, RBLOCK])
tmp85 = tl.sum(tmp83, 1)[:, None]
tmp86 = 64.0
tmp87 = tmp43 / tmp86
tmp88 = -0.125
tmp89 = tmp87 * tmp88
tmp90 = tmp85 / tmp86
tmp91 = tmp90 * tmp88
tmp92 = tmp89 + tmp91
tl.debug_barrier()
tl.store(in_out_ptr0 + (tl.full([XBLOCK, 1], 0, tl.int32)), tmp92, None)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, arg1_1, arg2_1, arg3_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg1_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg2_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg3_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf1 = empty_strided_cuda((), (), torch.float32)
buf4 = buf1; del buf1 # reuse
# Topologically Sorted Source Nodes: [mul, add, pow_1, sub, exp, pow_2, sub_1, sum_1, mean, kl_node, mul_2, add_1, pow_3, sub_2, exp_1, pow_4, sub_3, sum_2, mean_1, kl_edge, kl], Original ATen: [aten.mul, aten.add, aten.pow, aten.sub, aten.exp, aten.sum, aten.mean]
stream0 = get_raw_stream(0)
triton_per_fused_add_exp_mean_mul_pow_sub_sum_0.run(buf4, arg0_1, arg1_1, arg2_1, arg3_1, 1, 64, grid=grid(1), stream=stream0)
del arg0_1
del arg1_1
del arg2_1
del arg3_1
return (buf4, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
arg1_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
arg2_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
arg3_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([arg0_1, arg1_1, arg2_1, arg3_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| from torch.nn import Module
import torch
from torch.nn.modules.module import Module
class kl_loss(Module):
def __init__(self, num_nodes, num_edges):
super(kl_loss, self).__init__()
self.num_nodes = num_nodes
self.num_edges = num_edges
def forward(self, z_node_log_std, z_node_mean, z_edge_log_std, z_edge_mean
):
kl_node = -(0.5 / self.num_nodes) * torch.mean(torch.sum(1 + 2 *
z_node_log_std - torch.pow(z_node_mean, 2) - torch.pow(torch.
exp(z_node_log_std), 2), 1))
kl_edge = -(0.5 / self.num_edges) * torch.mean(torch.sum(1 + 2 *
z_edge_log_std - torch.pow(z_edge_mean, 2) - torch.pow(torch.
exp(z_edge_log_std), 2), 1))
kl = kl_node + kl_edge
return kl
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4]), torch.rand(
[4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'num_nodes': 4, 'num_edges': 4}]
| import torch
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime.triton_helpers import math as tl_math
from torch.nn import Module
from torch.nn.modules.module import Module
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_per_fused_add_exp_mean_mul_pow_sub_sum_0(in_out_ptr0, in_ptr0,
in_ptr1, in_ptr2, in_ptr3, xnumel, rnumel, XBLOCK: tl.constexpr):
RBLOCK: tl.constexpr = 64
xoffset = tl.program_id(0) * XBLOCK
xoffset + tl.arange(0, XBLOCK)[:, None]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[None, :]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
r0 = rindex % 16
r1 = rindex // 16
tmp0 = tl.load(in_ptr0 + (r0 + 64 * r1), None)
tmp5 = tl.load(in_ptr1 + (r0 + 64 * r1), None)
tmp11 = tl.load(in_ptr0 + (16 + r0 + 64 * r1), None)
tmp14 = tl.load(in_ptr1 + (16 + r0 + 64 * r1), None)
tmp21 = tl.load(in_ptr0 + (32 + r0 + 64 * r1), None)
tmp24 = tl.load(in_ptr1 + (32 + r0 + 64 * r1), None)
tmp31 = tl.load(in_ptr0 + (48 + r0 + 64 * r1), None)
tmp34 = tl.load(in_ptr1 + (48 + r0 + 64 * r1), None)
tmp44 = tl.load(in_ptr2 + (r0 + 64 * r1), None)
tmp47 = tl.load(in_ptr3 + (r0 + 64 * r1), None)
tmp53 = tl.load(in_ptr2 + (16 + r0 + 64 * r1), None)
tmp56 = tl.load(in_ptr3 + (16 + r0 + 64 * r1), None)
tmp63 = tl.load(in_ptr2 + (32 + r0 + 64 * r1), None)
tmp66 = tl.load(in_ptr3 + (32 + r0 + 64 * r1), None)
tmp73 = tl.load(in_ptr2 + (48 + r0 + 64 * r1), None)
tmp76 = tl.load(in_ptr3 + (48 + r0 + 64 * r1), None)
tmp1 = 2.0
tmp2 = tmp0 * tmp1
tmp3 = 1.0
tmp4 = tmp2 + tmp3
tmp6 = tmp5 * tmp5
tmp7 = tmp4 - tmp6
tmp8 = tl_math.exp(tmp0)
tmp9 = tmp8 * tmp8
tmp10 = tmp7 - tmp9
tmp12 = tmp11 * tmp1
tmp13 = tmp12 + tmp3
tmp15 = tmp14 * tmp14
tmp16 = tmp13 - tmp15
tmp17 = tl_math.exp(tmp11)
tmp18 = tmp17 * tmp17
tmp19 = tmp16 - tmp18
tmp20 = tmp10 + tmp19
tmp22 = tmp21 * tmp1
tmp23 = tmp22 + tmp3
tmp25 = tmp24 * tmp24
tmp26 = tmp23 - tmp25
tmp27 = tl_math.exp(tmp21)
tmp28 = tmp27 * tmp27
tmp29 = tmp26 - tmp28
tmp30 = tmp20 + tmp29
tmp32 = tmp31 * tmp1
tmp33 = tmp32 + tmp3
tmp35 = tmp34 * tmp34
tmp36 = tmp33 - tmp35
tmp37 = tl_math.exp(tmp31)
tmp38 = tmp37 * tmp37
tmp39 = tmp36 - tmp38
tmp40 = tmp30 + tmp39
tmp41 = tl.broadcast_to(tmp40, [XBLOCK, RBLOCK])
tmp43 = tl.sum(tmp41, 1)[:, None]
tmp45 = tmp44 * tmp1
tmp46 = tmp45 + tmp3
tmp48 = tmp47 * tmp47
tmp49 = tmp46 - tmp48
tmp50 = tl_math.exp(tmp44)
tmp51 = tmp50 * tmp50
tmp52 = tmp49 - tmp51
tmp54 = tmp53 * tmp1
tmp55 = tmp54 + tmp3
tmp57 = tmp56 * tmp56
tmp58 = tmp55 - tmp57
tmp59 = tl_math.exp(tmp53)
tmp60 = tmp59 * tmp59
tmp61 = tmp58 - tmp60
tmp62 = tmp52 + tmp61
tmp64 = tmp63 * tmp1
tmp65 = tmp64 + tmp3
tmp67 = tmp66 * tmp66
tmp68 = tmp65 - tmp67
tmp69 = tl_math.exp(tmp63)
tmp70 = tmp69 * tmp69
tmp71 = tmp68 - tmp70
tmp72 = tmp62 + tmp71
tmp74 = tmp73 * tmp1
tmp75 = tmp74 + tmp3
tmp77 = tmp76 * tmp76
tmp78 = tmp75 - tmp77
tmp79 = tl_math.exp(tmp73)
tmp80 = tmp79 * tmp79
tmp81 = tmp78 - tmp80
tmp82 = tmp72 + tmp81
tmp83 = tl.broadcast_to(tmp82, [XBLOCK, RBLOCK])
tmp85 = tl.sum(tmp83, 1)[:, None]
tmp86 = 64.0
tmp87 = tmp43 / tmp86
tmp88 = -0.125
tmp89 = tmp87 * tmp88
tmp90 = tmp85 / tmp86
tmp91 = tmp90 * tmp88
tmp92 = tmp89 + tmp91
tl.debug_barrier()
tl.store(in_out_ptr0 + tl.full([XBLOCK, 1], 0, tl.int32), tmp92, None)
def call(args):
arg0_1, arg1_1, arg2_1, arg3_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg1_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg2_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg3_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf1 = empty_strided_cuda((), (), torch.float32)
buf4 = buf1
del buf1
get_raw_stream(0)
triton_per_fused_add_exp_mean_mul_pow_sub_sum_0[grid(1)](buf4,
arg0_1, arg1_1, arg2_1, arg3_1, 1, 64, XBLOCK=1, num_warps=2,
num_stages=1)
del arg0_1
del arg1_1
del arg2_1
del arg3_1
return buf4,
class kl_lossNew(Module):
def __init__(self, num_nodes, num_edges):
super(kl_lossNew, self).__init__()
self.num_nodes = num_nodes
self.num_edges = num_edges
def forward(self, input_0, input_1, input_2, input_3):
arg0_1 = input_0
arg1_1 = input_1
arg2_1 = input_2
arg3_1 = input_3
output = call([arg0_1, arg1_1, arg2_1, arg3_1])
return output[0]
| iMoonLab/HHDTI | kl_loss | false | 6,843 | [
"MIT"
] | 1 | b2dd0e78818888e676afc91af1425dada5b3258a | https://github.com/iMoonLab/HHDTI/tree/b2dd0e78818888e676afc91af1425dada5b3258a | from torch.nn import Module
import torch
from torch.nn.modules.module import Module
class Model(Module):
def __init__(self, num_nodes, num_edges):
super().__init__()
self.num_nodes = num_nodes
self.num_edges = num_edges
def forward(self, z_node_log_std, z_node_mean, z_edge_log_std, z_edge_mean
):
kl_node = -(0.5 / self.num_nodes) * torch.mean(torch.sum(1 + 2 *
z_node_log_std - torch.pow(z_node_mean, 2) - torch.pow(torch.
exp(z_node_log_std), 2), 1))
kl_edge = -(0.5 / self.num_edges) * torch.mean(torch.sum(1 + 2 *
z_edge_log_std - torch.pow(z_edge_mean, 2) - torch.pow(torch.
exp(z_edge_log_std), 2), 1))
kl = kl_node + kl_edge
return kl
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4]), torch.rand(
[4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4, 4]
|
HGNN_conv | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/v6/cv6odvhmmcyvquog4eo62pdliew53orxzwe2wfzampr64jy3ppa7.py
# Topologically Sorted Source Nodes: [x_1], Original ATen: [aten.add]
# Source node to ATen node mapping:
# x_1 => add
# Graph fragment:
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%view_1, %primals_3), kwargs = {})
triton_poi_fused_add_0 = async_compile.triton('triton_poi_fused_add_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_0(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + (x2), tmp2, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (4, 4), (4, 1))
assert_size_stride(primals_3, (4, ), (1, ))
assert_size_stride(primals_4, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.mm]
extern_kernels.mm(reinterpret_tensor(primals_1, (64, 4), (4, 1), 0), primals_2, out=buf0)
del primals_2
buf1 = reinterpret_tensor(buf0, (4, 4, 4, 4), (64, 16, 4, 1), 0); del buf0 # reuse
# Topologically Sorted Source Nodes: [x_1], Original ATen: [aten.add]
stream0 = get_raw_stream(0)
triton_poi_fused_add_0.run(buf1, primals_3, 256, grid=grid(256), stream=stream0)
del primals_3
buf2 = empty_strided_cuda((16, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [x_2], Original ATen: [aten.bmm]
extern_kernels.bmm(reinterpret_tensor(primals_4, (16, 4, 4), (16, 4, 1), 0), reinterpret_tensor(buf1, (16, 4, 4), (16, 4, 1), 0), out=buf2)
del buf1
return (reinterpret_tensor(buf2, (4, 4, 4, 4), (64, 16, 4, 1), 0), reinterpret_tensor(primals_4, (16, 4, 4), (16, 1, 4), 0), reinterpret_tensor(primals_1, (4, 64), (1, 4), 0), )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import math
import torch
import torch.nn as nn
from torch.nn.parameter import Parameter
class HGNN_conv(nn.Module):
def __init__(self, in_ft, out_ft, bias=True):
super(HGNN_conv, self).__init__()
self.weight = Parameter(torch.Tensor(in_ft, out_ft))
if bias:
self.bias = Parameter(torch.Tensor(out_ft))
else:
self.register_parameter('bias', None)
self.reset_parameters()
def reset_parameters(self):
stdv = 1.0 / math.sqrt(self.weight.size(1))
self.weight.data.uniform_(-stdv, stdv)
if self.bias is not None:
self.bias.data.uniform_(-stdv, stdv)
def forward(self, x, G):
x = x.matmul(self.weight)
if self.bias is not None:
x = x + self.bias
x = G.matmul(x)
return x
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'in_ft': 4, 'out_ft': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
import math
import torch.nn as nn
from torch.nn.parameter import Parameter
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_add_0(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + x2, tmp2, xmask)
def call(args):
primals_1, primals_2, primals_3, primals_4 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (4, 4), (4, 1))
assert_size_stride(primals_3, (4,), (1,))
assert_size_stride(primals_4, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_1, (64, 4), (4, 1), 0),
primals_2, out=buf0)
del primals_2
buf1 = reinterpret_tensor(buf0, (4, 4, 4, 4), (64, 16, 4, 1), 0)
del buf0
get_raw_stream(0)
triton_poi_fused_add_0[grid(256)](buf1, primals_3, 256, XBLOCK=128,
num_warps=4, num_stages=1)
del primals_3
buf2 = empty_strided_cuda((16, 4, 4), (16, 4, 1), torch.float32)
extern_kernels.bmm(reinterpret_tensor(primals_4, (16, 4, 4), (16, 4,
1), 0), reinterpret_tensor(buf1, (16, 4, 4), (16, 4, 1), 0),
out=buf2)
del buf1
return reinterpret_tensor(buf2, (4, 4, 4, 4), (64, 16, 4, 1), 0
), reinterpret_tensor(primals_4, (16, 4, 4), (16, 1, 4), 0
), reinterpret_tensor(primals_1, (4, 64), (1, 4), 0)
class HGNN_convNew(nn.Module):
def __init__(self, in_ft, out_ft, bias=True):
super(HGNN_convNew, self).__init__()
self.weight = Parameter(torch.Tensor(in_ft, out_ft))
if bias:
self.bias = Parameter(torch.Tensor(out_ft))
else:
self.register_parameter('bias', None)
self.reset_parameters()
def reset_parameters(self):
stdv = 1.0 / math.sqrt(self.weight.size(1))
self.weight.data.uniform_(-stdv, stdv)
if self.bias is not None:
self.bias.data.uniform_(-stdv, stdv)
def forward(self, input_0, input_1):
primals_2 = self.weight
primals_3 = self.bias
primals_1 = input_0
primals_4 = input_1
output = call([primals_1, primals_2, primals_3, primals_4])
return output[0]
| iMoonLab/HHDTI | HGNN_conv | false | 6,844 | [
"MIT"
] | 1 | b2dd0e78818888e676afc91af1425dada5b3258a | https://github.com/iMoonLab/HHDTI/tree/b2dd0e78818888e676afc91af1425dada5b3258a | import math
import torch
import torch.nn as nn
from torch.nn.parameter import Parameter
class Model(nn.Module):
def __init__(self, in_ft, out_ft, bias=True):
super().__init__()
self.weight = Parameter(torch.Tensor(in_ft, out_ft))
if bias:
self.bias = Parameter(torch.Tensor(out_ft))
else:
self.register_parameter('bias', None)
self.reset_parameters()
def reset_parameters(self):
stdv = 1.0 / math.sqrt(self.weight.size(1))
self.weight.data.uniform_(-stdv, stdv)
if self.bias is not None:
self.bias.data.uniform_(-stdv, stdv)
def forward(self, x, G):
x = x.matmul(self.weight)
if self.bias is not None:
x = x + self.bias
x = G.matmul(x)
return x
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4, 4]
|
ArcFace | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/cv/ccvh5pcjc57oskvep5v2okylauufvnesitt2j6dbyvr6fh3vwuer.py
# Topologically Sorted Source Nodes: [output], Original ATen: [aten.mul]
# Source node to ATen node mapping:
# output => mul_2
# Graph fragment:
# %mul_2 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%arg0_1, 1.0), kwargs = {})
triton_poi_fused_mul_0 = async_compile.triton('triton_poi_fused_mul_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_mul_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_mul_0(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (x0), xmask)
tmp1 = 1.0
tmp2 = tmp0 * tmp1
tl.store(out_ptr0 + (x0), tmp2, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/fz/cfzxcyhkmrsdm5u4bef4furnodily6f7e3cway2qcmgx7glijwgc.py
# Topologically Sorted Source Nodes: [output, gt, mul, pow_1, sub, clamp, sine, mul_1, phi, sub_2, phi_1, getitem, setitem], Original ATen: [aten.mul, aten.gt, aten.pow, aten.rsub, aten.clamp, aten.sqrt, aten.sub, aten.where, aten.index, aten.index_put]
# Source node to ATen node mapping:
# clamp => clamp_max, clamp_min
# getitem => index
# gt => gt
# mul => mul
# mul_1 => mul_1
# output => mul_2
# phi => sub_1
# phi_1 => where
# pow_1 => pow_1
# setitem => index_put
# sine => sqrt
# sub => sub
# sub_2 => sub_2
# Graph fragment:
# %mul_2 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%arg0_1, 1.0), kwargs = {})
# %gt : [num_users=1] = call_function[target=torch.ops.aten.gt.Scalar](args = (%arg0_1, -0.8775825618903726), kwargs = {})
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%arg0_1, 0.8775825618903728), kwargs = {})
# %pow_1 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%arg0_1, 2), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (1.0, %pow_1), kwargs = {})
# %clamp_min : [num_users=1] = call_function[target=torch.ops.aten.clamp_min.default](args = (%sub, 0), kwargs = {})
# %clamp_max : [num_users=1] = call_function[target=torch.ops.aten.clamp_max.default](args = (%clamp_min, 1), kwargs = {})
# %sqrt : [num_users=1] = call_function[target=torch.ops.aten.sqrt.default](args = (%clamp_max,), kwargs = {})
# %mul_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sqrt, 0.479425538604203), kwargs = {})
# %sub_1 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%mul, %mul_1), kwargs = {})
# %sub_2 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%arg0_1, 0.23971276930210156), kwargs = {})
# %where : [num_users=1] = call_function[target=torch.ops.aten.where.self](args = (%gt, %sub_1, %sub_2), kwargs = {})
# %index : [num_users=1] = call_function[target=torch.ops.aten.index.Tensor](args = (%where, [%lift_fresh_copy, %arg1_1]), kwargs = {})
# %index_put : [num_users=1] = call_function[target=torch.ops.aten.index_put_.default](args = (%mul_2, [%lift_fresh_copy_1, %arg1_1], %index), kwargs = {})
triton_poi_fused_clamp_gt_index_index_put_mul_pow_rsub_sqrt_sub_where_1 = async_compile.triton('triton_poi_fused_clamp_gt_index_index_put_mul_pow_rsub_sqrt_sub_where_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*i64', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clamp_gt_index_index_put_mul_pow_rsub_sqrt_sub_where_1', 'mutated_arg_names': ['out_ptr1'], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clamp_gt_index_index_put_mul_pow_rsub_sqrt_sub_where_1(in_ptr0, in_ptr1, out_ptr1, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = (xindex // 16)
x0 = xindex % 16
x2 = xindex
tmp11 = tl.load(in_ptr0 + (x1), xmask, eviction_policy='evict_last')
tmp0 = x1
tmp1 = tl.full([1], 2, tl.int64)
tmp2 = tmp0 < tmp1
tmp3 = tl.full([1], 1, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = tl.full([1], 0, tl.int64)
tmp6 = tl.where(tmp4, tmp5, tmp3)
tmp7 = tl.full([1], 3, tl.int64)
tmp8 = tmp0 < tmp7
tmp9 = tl.where(tmp8, tmp1, tmp7)
tmp10 = tl.where(tmp2, tmp6, tmp9)
tmp12 = tl.full([XBLOCK], 4, tl.int32)
tmp13 = tmp11 + tmp12
tmp14 = tmp11 < 0
tmp15 = tl.where(tmp14, tmp13, tmp11)
tl.device_assert(((0 <= tmp15) & (tmp15 < 4)) | ~(xmask), "index out of bounds: 0 <= tmp15 < 4")
tmp17 = tl.load(in_ptr1 + (x0 + (16*tmp15) + (64*tmp10)), xmask)
tmp18 = -0.8775825618903726
tmp19 = tmp17 > tmp18
tmp20 = 0.8775825618903728
tmp21 = tmp17 * tmp20
tmp22 = tmp17 * tmp17
tmp23 = 1.0
tmp24 = tmp23 - tmp22
tmp25 = 0.0
tmp26 = triton_helpers.maximum(tmp24, tmp25)
tmp27 = triton_helpers.minimum(tmp26, tmp23)
tmp28 = libdevice.sqrt(tmp27)
tmp29 = 0.479425538604203
tmp30 = tmp28 * tmp29
tmp31 = tmp21 - tmp30
tmp32 = 0.23971276930210156
tmp33 = tmp17 - tmp32
tmp34 = tl.where(tmp19, tmp31, tmp33)
tl.store(out_ptr1 + (x0 + (16*tmp15) + (64*tmp10)), tmp34, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/sj/csjvrzgejwjlion5dslc3ka7v7ir7aoe2ycuonqyyicqmbkmudnz.py
# Topologically Sorted Source Nodes: [mul_3], Original ATen: [aten.mul]
# Source node to ATen node mapping:
# mul_3 => mul_3
# Graph fragment:
# %mul_3 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%index_put, 64.0), kwargs = {})
triton_poi_fused_mul_2 = async_compile.triton('triton_poi_fused_mul_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_mul_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_mul_2(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (x0), xmask)
tmp1 = 64.0
tmp2 = tmp0 * tmp1
tl.store(out_ptr0 + (x0), tmp2, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, arg1_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg1_1, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf1 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [output], Original ATen: [aten.mul]
stream0 = get_raw_stream(0)
triton_poi_fused_mul_0.run(arg0_1, buf1, 256, grid=grid(256), stream=stream0)
# Topologically Sorted Source Nodes: [output, gt, mul, pow_1, sub, clamp, sine, mul_1, phi, sub_2, phi_1, getitem, setitem], Original ATen: [aten.mul, aten.gt, aten.pow, aten.rsub, aten.clamp, aten.sqrt, aten.sub, aten.where, aten.index, aten.index_put]
triton_poi_fused_clamp_gt_index_index_put_mul_pow_rsub_sqrt_sub_where_1.run(arg1_1, arg0_1, buf1, 64, grid=grid(64), stream=stream0)
del arg0_1
del arg1_1
buf3 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [mul_3], Original ATen: [aten.mul]
triton_poi_fused_mul_2.run(buf1, buf3, 256, grid=grid(256), stream=stream0)
del buf1
return (buf3, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
arg1_1 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.int64)
fn = lambda: call([arg0_1, arg1_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import math
import torch
from itertools import product as product
import torch.nn as nn
import torch.utils.data.distributed
class ArcFace(nn.Module):
def __init__(self, s=64.0, m=0.5):
"""ArcFace formula:
cos(m + theta) = cos(m)cos(theta) - sin(m)sin(theta)
Note that:
0 <= m + theta <= Pi
So if (m + theta) >= Pi, then theta >= Pi - m. In [0, Pi]
we have:
cos(theta) < cos(Pi - m)
So we can use cos(Pi - m) as threshold to check whether
(m + theta) go out of [0, Pi]
Args:
embedding_size: usually 128, 256, 512 ...
class_num: num of people when training
s: scale, see normface https://arxiv.org/abs/1704.06369
m: margin, see SphereFace, CosFace, and ArcFace paper
https://github.com/siriusdemon/Build-Your-Own-Face-Model/blob/master/recognition/model/metric.py
"""
super().__init__()
self.s = s
self.m = m
self.cos_m = math.cos(m)
self.sin_m = math.sin(m)
self.th = math.cos(math.pi - m)
self.mm = math.sin(math.pi - m) * m
def forward(self, cosine, label):
sine = (1.0 - cosine.pow(2)).clamp(0, 1).sqrt()
phi = cosine * self.cos_m - sine * self.sin_m
phi = torch.where(cosine > self.th, phi, cosine - self.mm)
output = cosine * 1.0
batch_size = len(output)
output[range(batch_size), label] = phi[range(batch_size), label]
return output * self.s
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.ones([4], dtype=torch.int64)]
def get_init_inputs():
return [[], {}]
| import torch
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import libdevice
import math
from itertools import product as product
import torch.nn as nn
import torch.utils.data.distributed
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_poi_fused_mul_0(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + x0, xmask)
tmp1 = 1.0
tmp2 = tmp0 * tmp1
tl.store(out_ptr0 + x0, tmp2, xmask)
@triton.jit
def triton_poi_fused_clamp_gt_index_index_put_mul_pow_rsub_sqrt_sub_where_1(
in_ptr0, in_ptr1, out_ptr1, xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = xindex // 16
x0 = xindex % 16
tmp11 = tl.load(in_ptr0 + x1, xmask, eviction_policy='evict_last')
tmp0 = x1
tmp1 = tl.full([1], 2, tl.int64)
tmp2 = tmp0 < tmp1
tmp3 = tl.full([1], 1, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = tl.full([1], 0, tl.int64)
tmp6 = tl.where(tmp4, tmp5, tmp3)
tmp7 = tl.full([1], 3, tl.int64)
tmp8 = tmp0 < tmp7
tmp9 = tl.where(tmp8, tmp1, tmp7)
tmp10 = tl.where(tmp2, tmp6, tmp9)
tmp12 = tl.full([XBLOCK], 4, tl.int32)
tmp13 = tmp11 + tmp12
tmp14 = tmp11 < 0
tmp15 = tl.where(tmp14, tmp13, tmp11)
tl.device_assert((0 <= tmp15) & (tmp15 < 4) | ~xmask,
'index out of bounds: 0 <= tmp15 < 4')
tmp17 = tl.load(in_ptr1 + (x0 + 16 * tmp15 + 64 * tmp10), xmask)
tmp18 = -0.8775825618903726
tmp19 = tmp17 > tmp18
tmp20 = 0.8775825618903728
tmp21 = tmp17 * tmp20
tmp22 = tmp17 * tmp17
tmp23 = 1.0
tmp24 = tmp23 - tmp22
tmp25 = 0.0
tmp26 = triton_helpers.maximum(tmp24, tmp25)
tmp27 = triton_helpers.minimum(tmp26, tmp23)
tmp28 = libdevice.sqrt(tmp27)
tmp29 = 0.479425538604203
tmp30 = tmp28 * tmp29
tmp31 = tmp21 - tmp30
tmp32 = 0.23971276930210156
tmp33 = tmp17 - tmp32
tmp34 = tl.where(tmp19, tmp31, tmp33)
tl.store(out_ptr1 + (x0 + 16 * tmp15 + 64 * tmp10), tmp34, xmask)
@triton.jit
def triton_poi_fused_mul_2(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + x0, xmask)
tmp1 = 64.0
tmp2 = tmp0 * tmp1
tl.store(out_ptr0 + x0, tmp2, xmask)
def call(args):
arg0_1, arg1_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg1_1, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf1 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_mul_0[grid(256)](arg0_1, buf1, 256, XBLOCK=256,
num_warps=4, num_stages=1)
triton_poi_fused_clamp_gt_index_index_put_mul_pow_rsub_sqrt_sub_where_1[
grid(64)](arg1_1, arg0_1, buf1, 64, XBLOCK=64, num_warps=1,
num_stages=1)
del arg0_1
del arg1_1
buf3 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
triton_poi_fused_mul_2[grid(256)](buf1, buf3, 256, XBLOCK=256,
num_warps=4, num_stages=1)
del buf1
return buf3,
class ArcFaceNew(nn.Module):
def __init__(self, s=64.0, m=0.5):
"""ArcFace formula:
cos(m + theta) = cos(m)cos(theta) - sin(m)sin(theta)
Note that:
0 <= m + theta <= Pi
So if (m + theta) >= Pi, then theta >= Pi - m. In [0, Pi]
we have:
cos(theta) < cos(Pi - m)
So we can use cos(Pi - m) as threshold to check whether
(m + theta) go out of [0, Pi]
Args:
embedding_size: usually 128, 256, 512 ...
class_num: num of people when training
s: scale, see normface https://arxiv.org/abs/1704.06369
m: margin, see SphereFace, CosFace, and ArcFace paper
https://github.com/siriusdemon/Build-Your-Own-Face-Model/blob/master/recognition/model/metric.py
"""
super().__init__()
self.s = s
self.m = m
self.cos_m = math.cos(m)
self.sin_m = math.sin(m)
self.th = math.cos(math.pi - m)
self.mm = math.sin(math.pi - m) * m
def forward(self, input_0, input_1):
arg0_1 = input_0
arg1_1 = input_1
output = call([arg0_1, arg1_1])
return output[0]
| iChenning/face_project | ArcFace | false | 6,845 | [
"MIT"
] | 1 | 8d70858817da4d15c7b513ae492034784f57f35f | https://github.com/iChenning/face_project/tree/8d70858817da4d15c7b513ae492034784f57f35f | import math
import torch
from itertools import product as product
import torch.nn as nn
import torch.utils.data.distributed
class Model(nn.Module):
def __init__(self, s=64.0, m=0.5):
"""ArcFace formula:
cos(m + theta) = cos(m)cos(theta) - sin(m)sin(theta)
Note that:
0 <= m + theta <= Pi
So if (m + theta) >= Pi, then theta >= Pi - m. In [0, Pi]
we have:
cos(theta) < cos(Pi - m)
So we can use cos(Pi - m) as threshold to check whether
(m + theta) go out of [0, Pi]
Args:
embedding_size: usually 128, 256, 512 ...
class_num: num of people when training
s: scale, see normface https://arxiv.org/abs/1704.06369
m: margin, see SphereFace, CosFace, and ArcFace paper
https://github.com/siriusdemon/Build-Your-Own-Face-Model/blob/master/recognition/model/metric.py
"""
super().__init__()
self.s = s
self.m = m
self.cos_m = math.cos(m)
self.sin_m = math.sin(m)
self.th = math.cos(math.pi - m)
self.mm = math.sin(math.pi - m) * m
def forward(self, cosine, label):
sine = (1.0 - cosine.pow(2)).clamp(0, 1).sqrt()
phi = cosine * self.cos_m - sine * self.sin_m
phi = torch.where(cosine > self.th, phi, cosine - self.mm)
output = cosine * 1.0
batch_size = len(output)
output[range(batch_size), label] = phi[range(batch_size), label]
return output * self.s
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.ones([4], dtype=torch.int64)]
def get_init_inputs():
return []
|
Model | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/hu/chuveqed25us2fnst2i5cron3ljd4d3wbp76lsj5pwofil4axuyo.py
# Topologically Sorted Source Nodes: [sub, div], Original ATen: [aten.sub, aten.div]
# Source node to ATen node mapping:
# div => div
# sub => sub
# Graph fragment:
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%arg1_1, %arg0_1), kwargs = {})
# %div : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%sub, %arg2_1), kwargs = {})
triton_poi_fused_div_sub_0 = async_compile.triton('triton_poi_fused_div_sub_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_div_sub_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 3, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_div_sub_0(in_ptr0, in_ptr1, in_ptr2, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (x0), xmask)
tmp1 = tl.load(in_ptr1 + (x0), xmask)
tmp3 = tl.load(in_ptr2 + (x0), xmask)
tmp2 = tmp0 - tmp1
tmp4 = tmp2 / tmp3
tl.store(out_ptr0 + (x0), tmp4, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, arg1_1, arg2_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg1_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg2_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [sub, div], Original ATen: [aten.sub, aten.div]
stream0 = get_raw_stream(0)
triton_poi_fused_div_sub_0.run(arg1_1, arg0_1, arg2_1, buf0, 256, grid=grid(256), stream=stream0)
del arg0_1
del arg1_1
del arg2_1
return (buf0, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
arg1_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
arg2_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([arg0_1, arg1_1, arg2_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
from torch import nn
class Model(nn.Module):
def forward(self, img: 'torch.Tensor', scale: 'torch.Tensor', mean:
'torch.Tensor'):
return torch.div(torch.sub(img, mean), scale)
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4]), torch.rand(
[4, 4, 4, 4])]
def get_init_inputs():
return [[], {}]
| import torch
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch import nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_poi_fused_div_sub_0(in_ptr0, in_ptr1, in_ptr2, out_ptr0, xnumel,
XBLOCK: tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + x0, xmask)
tmp1 = tl.load(in_ptr1 + x0, xmask)
tmp3 = tl.load(in_ptr2 + x0, xmask)
tmp2 = tmp0 - tmp1
tmp4 = tmp2 / tmp3
tl.store(out_ptr0 + x0, tmp4, xmask)
def call(args):
arg0_1, arg1_1, arg2_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg1_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg2_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_div_sub_0[grid(256)](arg1_1, arg0_1, arg2_1, buf0,
256, XBLOCK=256, num_warps=4, num_stages=1)
del arg0_1
del arg1_1
del arg2_1
return buf0,
class ModelNew(nn.Module):
def forward(self, input_0, input_1, input_2):
arg0_1 = input_0
arg1_1 = input_1
arg2_1 = input_2
output = call([arg0_1, arg1_1, arg2_1])
return output[0]
| ibaiGorordo/depthai-experiments | Model | false | 6,846 | [
"MIT"
] | 1 | cde67e277120ddac815cbad6360695759cca900f | https://github.com/ibaiGorordo/depthai-experiments/tree/cde67e277120ddac815cbad6360695759cca900f | import torch
from torch import nn
class Model(nn.Module):
def forward(self, img: 'torch.Tensor', scale: 'torch.Tensor', mean:
'torch.Tensor'):
return torch.div(torch.sub(img, mean), scale)
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4]), torch.rand(
[4, 4, 4, 4])]
def get_init_inputs():
return []
|
Actor | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/nc/cncwsucylpsg2zmlivjfxu6vbd64ztxjndlsix2ysjtby3xohgk4.py
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.tanh]
# Source node to ATen node mapping:
# x => tanh
# Graph fragment:
# %tanh : [num_users=2] = call_function[target=torch.ops.aten.tanh.default](args = (%view_1,), kwargs = {})
triton_poi_fused_tanh_0 = async_compile.triton('triton_poi_fused_tanh_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_tanh_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_tanh_0(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = libdevice.tanh(tmp2)
tl.store(in_out_ptr0 + (x2), tmp3, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (4, 4), (4, 1))
assert_size_stride(primals_3, (4, ), (1, ))
assert_size_stride(primals_4, (4, 4), (4, 1))
assert_size_stride(primals_5, (4, ), (1, ))
assert_size_stride(primals_6, (4, 4), (4, 1))
assert_size_stride(primals_7, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(primals_1, (64, 4), (4, 1), 0), reinterpret_tensor(primals_2, (4, 4), (1, 4), 0), out=buf0)
del primals_2
buf1 = reinterpret_tensor(buf0, (4, 4, 4, 4), (64, 16, 4, 1), 0); del buf0 # reuse
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.tanh]
stream0 = get_raw_stream(0)
triton_poi_fused_tanh_0.run(buf1, primals_3, 256, grid=grid(256), stream=stream0)
del primals_3
buf2 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(buf1, (64, 4), (4, 1), 0), reinterpret_tensor(primals_4, (4, 4), (1, 4), 0), out=buf2)
buf3 = reinterpret_tensor(buf2, (4, 4, 4, 4), (64, 16, 4, 1), 0); del buf2 # reuse
# Topologically Sorted Source Nodes: [x_1], Original ATen: [aten.tanh]
triton_poi_fused_tanh_0.run(buf3, primals_5, 256, grid=grid(256), stream=stream0)
del primals_5
buf4 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(buf3, (64, 4), (4, 1), 0), reinterpret_tensor(primals_6, (4, 4), (1, 4), 0), out=buf4)
buf5 = reinterpret_tensor(buf4, (4, 4, 4, 4), (64, 16, 4, 1), 0); del buf4 # reuse
# Topologically Sorted Source Nodes: [mu], Original ATen: [aten.tanh]
triton_poi_fused_tanh_0.run(buf5, primals_7, 256, grid=grid(256), stream=stream0)
del primals_7
return (buf5, reinterpret_tensor(primals_1, (64, 4), (4, 1), 0), buf1, buf3, buf5, primals_6, primals_4, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn.functional as F
import torch.nn as nn
class Actor(nn.Module):
def __init__(self, hidden_size, num_inputs, action_space):
super(Actor, self).__init__()
self.action_space = action_space
num_outputs = action_space.shape[0]
self.linear1 = nn.Linear(num_inputs, hidden_size)
nn.init.normal_(self.linear1.weight, 0.0, 0.02)
self.linear2 = nn.Linear(hidden_size, hidden_size)
nn.init.normal_(self.linear2.weight, 0.0, 0.02)
self.mu = nn.Linear(hidden_size, num_outputs)
torch.nn.init.uniform_(self.mu.weight, a=-0.003, b=0.003)
def forward(self, inputs):
x = inputs
x = F.tanh(self.linear1(x))
x = F.tanh(self.linear2(x))
mu = F.tanh(self.mu(x))
return mu
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'hidden_size': 4, 'num_inputs': 4, 'action_space': torch.
rand([4, 4])}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime.triton_helpers import libdevice
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_tanh_0(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = libdevice.tanh(tmp2)
tl.store(in_out_ptr0 + x2, tmp3, xmask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7) = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (4, 4), (4, 1))
assert_size_stride(primals_3, (4,), (1,))
assert_size_stride(primals_4, (4, 4), (4, 1))
assert_size_stride(primals_5, (4,), (1,))
assert_size_stride(primals_6, (4, 4), (4, 1))
assert_size_stride(primals_7, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_1, (64, 4), (4, 1), 0),
reinterpret_tensor(primals_2, (4, 4), (1, 4), 0), out=buf0)
del primals_2
buf1 = reinterpret_tensor(buf0, (4, 4, 4, 4), (64, 16, 4, 1), 0)
del buf0
get_raw_stream(0)
triton_poi_fused_tanh_0[grid(256)](buf1, primals_3, 256, XBLOCK=128,
num_warps=4, num_stages=1)
del primals_3
buf2 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(buf1, (64, 4), (4, 1), 0),
reinterpret_tensor(primals_4, (4, 4), (1, 4), 0), out=buf2)
buf3 = reinterpret_tensor(buf2, (4, 4, 4, 4), (64, 16, 4, 1), 0)
del buf2
triton_poi_fused_tanh_0[grid(256)](buf3, primals_5, 256, XBLOCK=128,
num_warps=4, num_stages=1)
del primals_5
buf4 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(buf3, (64, 4), (4, 1), 0),
reinterpret_tensor(primals_6, (4, 4), (1, 4), 0), out=buf4)
buf5 = reinterpret_tensor(buf4, (4, 4, 4, 4), (64, 16, 4, 1), 0)
del buf4
triton_poi_fused_tanh_0[grid(256)](buf5, primals_7, 256, XBLOCK=128,
num_warps=4, num_stages=1)
del primals_7
return buf5, reinterpret_tensor(primals_1, (64, 4), (4, 1), 0
), buf1, buf3, buf5, primals_6, primals_4
class ActorNew(nn.Module):
def __init__(self, hidden_size, num_inputs, action_space):
super(ActorNew, self).__init__()
self.action_space = action_space
num_outputs = action_space.shape[0]
self.linear1 = nn.Linear(num_inputs, hidden_size)
nn.init.normal_(self.linear1.weight, 0.0, 0.02)
self.linear2 = nn.Linear(hidden_size, hidden_size)
nn.init.normal_(self.linear2.weight, 0.0, 0.02)
self.mu = nn.Linear(hidden_size, num_outputs)
torch.nn.init.uniform_(self.mu.weight, a=-0.003, b=0.003)
def forward(self, input_0):
primals_2 = self.linear1.weight
primals_3 = self.linear1.bias
primals_4 = self.linear2.weight
primals_5 = self.linear2.bias
primals_6 = self.mu.weight
primals_7 = self.mu.bias
primals_1 = input_0
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7])
return output[0]
| icml2019-anonymous-author/Action-Robust-Reinforcement-Learning | Actor | false | 6,847 | [
"MIT"
] | 1 | 03f0a1dd5f4a0fc5230c0ad0b41f63161bae862b | https://github.com/icml2019-anonymous-author/Action-Robust-Reinforcement-Learning/tree/03f0a1dd5f4a0fc5230c0ad0b41f63161bae862b | import torch
import torch.nn.functional as F
import torch.nn as nn
class Model(nn.Module):
def __init__(self, hidden_size, num_inputs, action_space):
super().__init__()
self.action_space = action_space
num_outputs = action_space.shape[0]
self.linear1 = nn.Linear(num_inputs, hidden_size)
nn.init.normal_(self.linear1.weight, 0.0, 0.02)
self.linear2 = nn.Linear(hidden_size, hidden_size)
nn.init.normal_(self.linear2.weight, 0.0, 0.02)
self.mu = nn.Linear(hidden_size, num_outputs)
torch.nn.init.uniform_(self.mu.weight, a=-0.003, b=0.003)
def forward(self, inputs):
x = inputs
x = F.tanh(self.linear1(x))
x = F.tanh(self.linear2(x))
mu = F.tanh(self.mu(x))
return mu
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'hidden_size': 4, 'num_inputs': 4, 'action_space': torch.
rand([4, 4])}]
|
Block | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/wd/cwdz7kqs3uwyg53zsyekt77eye7yjl6v7vulow2q6ni534mkf6zw.py
# Topologically Sorted Source Nodes: [layer_norm], Original ATen: [aten.native_layer_norm]
# Source node to ATen node mapping:
# layer_norm => add, rsqrt, var_mean
# Graph fragment:
# %var_mean : [num_users=2] = call_function[target=torch.ops.aten.var_mean.correction](args = (%primals_3, [2]), kwargs = {correction: 0, keepdim: True})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%getitem, 1e-05), kwargs = {})
# %rsqrt : [num_users=1] = call_function[target=torch.ops.aten.rsqrt.default](args = (%add,), kwargs = {})
triton_poi_fused_native_layer_norm_0 = async_compile.triton('triton_poi_fused_native_layer_norm_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_native_layer_norm_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 4, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_native_layer_norm_0(in_ptr0, out_ptr0, out_ptr1, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (4*x0), xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr0 + (1 + (4*x0)), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr0 + (2 + (4*x0)), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (3 + (4*x0)), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp4 = tmp2 + tmp3
tmp6 = tmp4 + tmp5
tmp7 = 4.0
tmp8 = tmp6 / tmp7
tmp9 = tmp0 - tmp8
tmp10 = tmp9 * tmp9
tmp11 = tmp1 - tmp8
tmp12 = tmp11 * tmp11
tmp13 = tmp10 + tmp12
tmp14 = tmp3 - tmp8
tmp15 = tmp14 * tmp14
tmp16 = tmp13 + tmp15
tmp17 = tmp5 - tmp8
tmp18 = tmp17 * tmp17
tmp19 = tmp16 + tmp18
tmp20 = tmp19 / tmp7
tmp21 = 1e-05
tmp22 = tmp20 + tmp21
tmp23 = libdevice.rsqrt(tmp22)
tl.store(out_ptr0 + (x0), tmp8, xmask)
tl.store(out_ptr1 + (x0), tmp23, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/vs/cvsfvbs4wlaqvwxm3svg65dnhcq336ptudvn6xetnbnrtzj7xssn.py
# Topologically Sorted Source Nodes: [layer_norm], Original ATen: [aten.native_layer_norm]
# Source node to ATen node mapping:
# layer_norm => add, add_1, mul, mul_1, rsqrt, sub, var_mean
# Graph fragment:
# %var_mean : [num_users=2] = call_function[target=torch.ops.aten.var_mean.correction](args = (%primals_3, [2]), kwargs = {correction: 0, keepdim: True})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%getitem, 1e-05), kwargs = {})
# %rsqrt : [num_users=1] = call_function[target=torch.ops.aten.rsqrt.default](args = (%add,), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%primals_3, %getitem_1), kwargs = {})
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sub, %rsqrt), kwargs = {})
# %mul_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%mul, %primals_1), kwargs = {})
# %add_1 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul_1, %primals_2), kwargs = {})
triton_poi_fused_native_layer_norm_1 = async_compile.triton('triton_poi_fused_native_layer_norm_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: '*fp32', 5: '*fp32', 6: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5, 6), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_native_layer_norm_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_native_layer_norm_1(in_ptr0, in_ptr1, in_ptr2, in_ptr3, in_ptr4, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
x0 = xindex % 4
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr1 + (x1), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr2 + (x1), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr3 + (x0), xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr4 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 - tmp1
tmp4 = tmp2 * tmp3
tmp6 = tmp4 * tmp5
tmp8 = tmp6 + tmp7
tl.store(out_ptr0 + (x2), tmp8, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/vs/cvs6icdxviah4kvxj7x53zy2cxi5vohxx6kljkphckzrvc3cntr4.py
# Topologically Sorted Source Nodes: [q_1, attn], Original ATen: [aten.mul, aten.clone]
# Source node to ATen node mapping:
# attn => clone
# q_1 => mul_2
# Graph fragment:
# %mul_2 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%select, 1.0), kwargs = {})
# %clone : [num_users=1] = call_function[target=torch.ops.aten.clone.default](args = (%expand,), kwargs = {memory_format: torch.contiguous_format})
triton_poi_fused_clone_mul_2 = async_compile.triton('triton_poi_fused_clone_mul_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16, 4], tile_hint=TileHint.SQUARE,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clone_mul_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clone_mul_2(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = (yindex // 4)
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + (12*x2) + (48*y1)), xmask & ymask, eviction_policy='evict_last')
tmp1 = 1.0
tmp2 = tmp0 * tmp1
tl.store(out_ptr0 + (x2 + (4*y3)), tmp2, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/aw/cawvwx3nv7ipnpnf2hcgwz5usu7vsw5yynj5ofrunhktjwqff5vq.py
# Topologically Sorted Source Nodes: [attn], Original ATen: [aten.clone]
# Source node to ATen node mapping:
# attn => clone_1
# Graph fragment:
# %clone_1 : [num_users=1] = call_function[target=torch.ops.aten.clone.default](args = (%expand_1,), kwargs = {memory_format: torch.contiguous_format})
triton_poi_fused_clone_3 = async_compile.triton('triton_poi_fused_clone_3', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16, 4], tile_hint=TileHint.SQUARE,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clone_3', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clone_3(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = (yindex // 4)
y3 = yindex
tmp0 = tl.load(in_ptr0 + (4 + y0 + (12*x2) + (48*y1)), xmask & ymask, eviction_policy='evict_last')
tl.store(out_ptr0 + (x2 + (4*y3)), tmp0, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/iv/civdgwzpphyda4rs4fr3g6w25bprv7bn4anqgivrgzavi7xr5pdl.py
# Topologically Sorted Source Nodes: [attn_1], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# attn_1 => amax, exp, sub_1
# Graph fragment:
# %amax : [num_users=1] = call_function[target=torch.ops.aten.amax.default](args = (%view_5, [-1], True), kwargs = {})
# %sub_1 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%view_5, %amax), kwargs = {})
# %exp : [num_users=2] = call_function[target=torch.ops.aten.exp.default](args = (%sub_1,), kwargs = {})
triton_poi_fused__softmax_4 = async_compile.triton('triton_poi_fused__softmax_4', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_4', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_4(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp3 = triton_helpers.maximum(tmp1, tmp2)
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp7 = triton_helpers.maximum(tmp5, tmp6)
tmp8 = tmp0 - tmp7
tmp9 = tl_math.exp(tmp8)
tl.store(out_ptr0 + (x2), tmp9, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/a4/ca4u6hbohfqkgchihihlu5hrf3vuqm27r2ncsg7xb6g4ikttl2at.py
# Topologically Sorted Source Nodes: [attn_1], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# attn_1 => div, sum_1
# Graph fragment:
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%exp, [-1], True), kwargs = {})
# %div : [num_users=2] = call_function[target=torch.ops.aten.div.Tensor](args = (%exp, %sum_1), kwargs = {})
triton_poi_fused__softmax_5 = async_compile.triton('triton_poi_fused__softmax_5', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_5', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_5(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + (x2), tmp8, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/vv/cvvhis67uzj3m3ebbd4sgghaemqhihabasphltk5wytqdd6fe74t.py
# Topologically Sorted Source Nodes: [matmul_1], Original ATen: [aten.clone]
# Source node to ATen node mapping:
# matmul_1 => clone_3
# Graph fragment:
# %clone_3 : [num_users=1] = call_function[target=torch.ops.aten.clone.default](args = (%expand_3,), kwargs = {memory_format: torch.contiguous_format})
triton_poi_fused_clone_6 = async_compile.triton('triton_poi_fused_clone_6', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16, 4], tile_hint=TileHint.SQUARE,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clone_6', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clone_6(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = (yindex // 4)
y3 = yindex
tmp0 = tl.load(in_ptr0 + (8 + y0 + (12*x2) + (48*y1)), xmask & ymask, eviction_policy='evict_last')
tl.store(out_ptr0 + (x2 + (4*y3)), tmp0, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/lw/clwfsjrjxeb2gmxy5p3lplvcrvrn37iuw4atjria32bxp2jajrtc.py
# Topologically Sorted Source Nodes: [x_1], Original ATen: [aten.clone]
# Source node to ATen node mapping:
# x_1 => clone_4
# Graph fragment:
# %clone_4 : [num_users=1] = call_function[target=torch.ops.aten.clone.default](args = (%view_9,), kwargs = {memory_format: torch.contiguous_format})
triton_poi_fused_clone_7 = async_compile.triton('triton_poi_fused_clone_7', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16, 4], tile_hint=TileHint.SQUARE,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clone_7', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clone_7(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = (yindex // 4)
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + (4*x2) + (16*y1)), xmask & ymask, eviction_policy='evict_last')
tl.store(out_ptr0 + (x2 + (4*y3)), tmp0, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/5y/c5yhyv7emyc7i2ozpvns6tsiqcvdzktqqpohy4sedfe7aihkojch.py
# Topologically Sorted Source Nodes: [x_1, x_3, layer_norm_1], Original ATen: [aten.add, aten.native_layer_norm]
# Source node to ATen node mapping:
# layer_norm_1 => var_mean_1
# x_1 => add_2
# x_3 => add_3
# Graph fragment:
# %add_2 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%view_11, %primals_6), kwargs = {})
# %add_3 : [num_users=3] = call_function[target=torch.ops.aten.add.Tensor](args = (%primals_3, %add_2), kwargs = {})
# %var_mean_1 : [num_users=2] = call_function[target=torch.ops.aten.var_mean.correction](args = (%add_3, [2]), kwargs = {correction: 0, keepdim: True})
triton_poi_fused_add_native_layer_norm_8 = async_compile.triton('triton_poi_fused_add_native_layer_norm_8', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: '*fp32', 5: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_native_layer_norm_8', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 12, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_native_layer_norm_8(in_ptr0, in_ptr1, in_ptr2, out_ptr0, out_ptr1, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (4*x0), xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (4*x0), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr2 + (0))
tmp3 = tl.broadcast_to(tmp2, [XBLOCK])
tmp6 = tl.load(in_ptr0 + (1 + (4*x0)), xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr1 + (1 + (4*x0)), xmask, eviction_policy='evict_last')
tmp8 = tl.load(in_ptr2 + (1))
tmp9 = tl.broadcast_to(tmp8, [XBLOCK])
tmp13 = tl.load(in_ptr0 + (2 + (4*x0)), xmask, eviction_policy='evict_last')
tmp14 = tl.load(in_ptr1 + (2 + (4*x0)), xmask, eviction_policy='evict_last')
tmp15 = tl.load(in_ptr2 + (2))
tmp16 = tl.broadcast_to(tmp15, [XBLOCK])
tmp20 = tl.load(in_ptr0 + (3 + (4*x0)), xmask, eviction_policy='evict_last')
tmp21 = tl.load(in_ptr1 + (3 + (4*x0)), xmask, eviction_policy='evict_last')
tmp22 = tl.load(in_ptr2 + (3))
tmp23 = tl.broadcast_to(tmp22, [XBLOCK])
tmp4 = tmp1 + tmp3
tmp5 = tmp0 + tmp4
tmp10 = tmp7 + tmp9
tmp11 = tmp6 + tmp10
tmp12 = tmp5 + tmp11
tmp17 = tmp14 + tmp16
tmp18 = tmp13 + tmp17
tmp19 = tmp12 + tmp18
tmp24 = tmp21 + tmp23
tmp25 = tmp20 + tmp24
tmp26 = tmp19 + tmp25
tmp27 = 4.0
tmp28 = tmp26 / tmp27
tmp29 = tmp5 - tmp28
tmp30 = tmp29 * tmp29
tmp31 = tmp11 - tmp28
tmp32 = tmp31 * tmp31
tmp33 = tmp30 + tmp32
tmp34 = tmp18 - tmp28
tmp35 = tmp34 * tmp34
tmp36 = tmp33 + tmp35
tmp37 = tmp25 - tmp28
tmp38 = tmp37 * tmp37
tmp39 = tmp36 + tmp38
tmp40 = tmp39 / tmp27
tl.store(out_ptr0 + (x0), tmp28, xmask)
tl.store(out_ptr1 + (x0), tmp40, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/xj/cxjpr2ute76xkk7edg7qlvolks2ggx2xwbrttteralhmvd2xsktw.py
# Topologically Sorted Source Nodes: [x_1, x_3, layer_norm_1], Original ATen: [aten.add, aten.native_layer_norm]
# Source node to ATen node mapping:
# layer_norm_1 => add_4, add_5, mul_3, mul_4, rsqrt_1, sub_2
# x_1 => add_2
# x_3 => add_3
# Graph fragment:
# %add_2 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%view_11, %primals_6), kwargs = {})
# %add_3 : [num_users=3] = call_function[target=torch.ops.aten.add.Tensor](args = (%primals_3, %add_2), kwargs = {})
# %add_4 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%getitem_2, 1e-05), kwargs = {})
# %rsqrt_1 : [num_users=1] = call_function[target=torch.ops.aten.rsqrt.default](args = (%add_4,), kwargs = {})
# %sub_2 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%add_3, %getitem_3), kwargs = {})
# %mul_3 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sub_2, %rsqrt_1), kwargs = {})
# %mul_4 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%mul_3, %primals_7), kwargs = {})
# %add_5 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul_4, %primals_8), kwargs = {})
triton_poi_fused_add_native_layer_norm_9 = async_compile.triton('triton_poi_fused_add_native_layer_norm_9', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: '*fp32', 5: '*fp32', 6: '*fp32', 7: '*fp32', 8: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5, 6, 7, 8), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_native_layer_norm_9', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 7, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_native_layer_norm_9(in_ptr0, in_ptr1, in_ptr2, in_ptr3, in_ptr4, in_ptr5, in_ptr6, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr1 + (x2), xmask)
tmp2 = tl.load(in_ptr2 + (x0), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr3 + (x1), xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr4 + (x1), xmask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr5 + (x0), xmask, eviction_policy='evict_last')
tmp14 = tl.load(in_ptr6 + (x0), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp4 = tmp0 + tmp3
tmp6 = tmp4 - tmp5
tmp8 = 1e-05
tmp9 = tmp7 + tmp8
tmp10 = libdevice.rsqrt(tmp9)
tmp11 = tmp6 * tmp10
tmp13 = tmp11 * tmp12
tmp15 = tmp13 + tmp14
tl.store(out_ptr0 + (x2), tmp15, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/b4/cb43jhxvcrefkhdp7ixdoh6nmvez5h55vhlzkxtasuovu5ru7pe5.py
# Topologically Sorted Source Nodes: [x_5], Original ATen: [aten.gelu]
# Source node to ATen node mapping:
# x_5 => add_6, erf, mul_5, mul_6, mul_7
# Graph fragment:
# %mul_5 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%view_13, 0.5), kwargs = {})
# %mul_6 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%view_13, 0.7071067811865476), kwargs = {})
# %erf : [num_users=1] = call_function[target=torch.ops.aten.erf.default](args = (%mul_6,), kwargs = {})
# %add_6 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%erf, 1), kwargs = {})
# %mul_7 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%mul_5, %add_6), kwargs = {})
triton_poi_fused_gelu_10 = async_compile.triton('triton_poi_fused_gelu_10', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_gelu_10', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_gelu_10(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (x0), xmask)
tmp1 = 0.5
tmp2 = tmp0 * tmp1
tmp3 = 0.7071067811865476
tmp4 = tmp0 * tmp3
tmp5 = libdevice.erf(tmp4)
tmp6 = 1.0
tmp7 = tmp5 + tmp6
tmp8 = tmp2 * tmp7
tl.store(out_ptr0 + (x0), tmp8, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/pu/cpuql3oz4hmaygynopg7lq7xhfiv7hr7pr4vyzhfpmw34jymdp7q.py
# Topologically Sorted Source Nodes: [x_1, x_3, x_8], Original ATen: [aten.add]
# Source node to ATen node mapping:
# x_1 => add_2
# x_3 => add_3
# x_8 => add_7
# Graph fragment:
# %add_2 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%view_11, %primals_6), kwargs = {})
# %add_3 : [num_users=3] = call_function[target=torch.ops.aten.add.Tensor](args = (%primals_3, %add_2), kwargs = {})
# %add_7 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%add_3, %view_15), kwargs = {})
triton_poi_fused_add_11 = async_compile.triton('triton_poi_fused_add_11', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: '*fp32', 5: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_11', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_11(in_out_ptr0, in_ptr0, in_ptr1, in_ptr2, in_ptr3, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr1 + (x2), xmask)
tmp2 = tl.load(in_ptr2 + (x0), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_out_ptr0 + (x2), xmask)
tmp6 = tl.load(in_ptr3 + (x0), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp4 = tmp0 + tmp3
tmp7 = tmp5 + tmp6
tmp8 = tmp4 + tmp7
tl.store(in_out_ptr0 + (x2), tmp8, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9, primals_10, primals_11, primals_12 = args
args.clear()
assert_size_stride(primals_1, (4, ), (1, ))
assert_size_stride(primals_2, (4, ), (1, ))
assert_size_stride(primals_3, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_4, (12, 4), (4, 1))
assert_size_stride(primals_5, (4, 4), (4, 1))
assert_size_stride(primals_6, (4, ), (1, ))
assert_size_stride(primals_7, (4, ), (1, ))
assert_size_stride(primals_8, (4, ), (1, ))
assert_size_stride(primals_9, (16, 4), (4, 1))
assert_size_stride(primals_10, (16, ), (1, ))
assert_size_stride(primals_11, (4, 16), (16, 1))
assert_size_stride(primals_12, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 1), (4, 1, 16), torch.float32)
buf1 = empty_strided_cuda((4, 4, 1), (4, 1, 16), torch.float32)
# Topologically Sorted Source Nodes: [layer_norm], Original ATen: [aten.native_layer_norm]
stream0 = get_raw_stream(0)
triton_poi_fused_native_layer_norm_0.run(primals_3, buf0, buf1, 16, grid=grid(16), stream=stream0)
buf2 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [layer_norm], Original ATen: [aten.native_layer_norm]
triton_poi_fused_native_layer_norm_1.run(primals_3, buf0, buf1, primals_1, primals_2, buf2, 64, grid=grid(64), stream=stream0)
del primals_1
del primals_2
buf3 = empty_strided_cuda((16, 12), (12, 1), torch.float32)
# Topologically Sorted Source Nodes: [qkv], Original ATen: [aten.mm]
extern_kernels.mm(reinterpret_tensor(buf2, (16, 4), (4, 1), 0), reinterpret_tensor(primals_4, (4, 12), (1, 4), 0), out=buf3)
buf4 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 1), torch.float32)
# Topologically Sorted Source Nodes: [q_1, attn], Original ATen: [aten.mul, aten.clone]
triton_poi_fused_clone_mul_2.run(buf3, buf4, 16, 4, grid=grid(16, 4), stream=stream0)
buf5 = empty_strided_cuda((4, 4, 1, 4), (16, 4, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [attn], Original ATen: [aten.clone]
triton_poi_fused_clone_3.run(buf3, buf5, 16, 4, grid=grid(16, 4), stream=stream0)
buf6 = empty_strided_cuda((16, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [attn], Original ATen: [aten.bmm]
extern_kernels.bmm(reinterpret_tensor(buf4, (16, 4, 1), (4, 1, 0), 0), reinterpret_tensor(buf5, (16, 1, 4), (4, 0, 1), 0), out=buf6)
buf7 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [attn_1], Original ATen: [aten._softmax]
triton_poi_fused__softmax_4.run(buf6, buf7, 256, grid=grid(256), stream=stream0)
buf8 = reinterpret_tensor(buf6, (4, 4, 4, 4), (64, 16, 4, 1), 0); del buf6 # reuse
# Topologically Sorted Source Nodes: [attn_1], Original ATen: [aten._softmax]
triton_poi_fused__softmax_5.run(buf7, buf8, 256, grid=grid(256), stream=stream0)
buf9 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 1), torch.float32)
# Topologically Sorted Source Nodes: [matmul_1], Original ATen: [aten.clone]
triton_poi_fused_clone_6.run(buf3, buf9, 16, 4, grid=grid(16, 4), stream=stream0)
del buf3
buf10 = empty_strided_cuda((16, 4, 1), (4, 1, 1), torch.float32)
# Topologically Sorted Source Nodes: [matmul_1], Original ATen: [aten.bmm]
extern_kernels.bmm(reinterpret_tensor(buf8, (16, 4, 4), (16, 4, 1), 0), reinterpret_tensor(buf9, (16, 4, 1), (4, 1, 0), 0), out=buf10)
buf11 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [x_1], Original ATen: [aten.clone]
triton_poi_fused_clone_7.run(buf10, buf11, 16, 4, grid=grid(16, 4), stream=stream0)
buf12 = reinterpret_tensor(buf10, (16, 4), (4, 1), 0); del buf10 # reuse
# Topologically Sorted Source Nodes: [x_1], Original ATen: [aten.mm]
extern_kernels.mm(reinterpret_tensor(buf11, (16, 4), (4, 1), 0), reinterpret_tensor(primals_5, (4, 4), (1, 4), 0), out=buf12)
buf13 = buf1; del buf1 # reuse
buf14 = buf0; del buf0 # reuse
# Topologically Sorted Source Nodes: [x_1, x_3, layer_norm_1], Original ATen: [aten.add, aten.native_layer_norm]
triton_poi_fused_add_native_layer_norm_8.run(primals_3, buf12, primals_6, buf13, buf14, 16, grid=grid(16), stream=stream0)
buf15 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [x_1, x_3, layer_norm_1], Original ATen: [aten.add, aten.native_layer_norm]
triton_poi_fused_add_native_layer_norm_9.run(primals_3, buf12, primals_6, buf13, buf14, primals_7, primals_8, buf15, 64, grid=grid(64), stream=stream0)
del buf13
del buf14
del primals_8
buf16 = reinterpret_tensor(buf7, (16, 16), (16, 1), 0); del buf7 # reuse
# Topologically Sorted Source Nodes: [x_4], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_10, reinterpret_tensor(buf15, (16, 4), (4, 1), 0), reinterpret_tensor(primals_9, (4, 16), (1, 4), 0), alpha=1, beta=1, out=buf16)
del primals_10
buf17 = empty_strided_cuda((4, 4, 16), (64, 16, 1), torch.float32)
# Topologically Sorted Source Nodes: [x_5], Original ATen: [aten.gelu]
triton_poi_fused_gelu_10.run(buf16, buf17, 256, grid=grid(256), stream=stream0)
buf18 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(buf17, (16, 16), (16, 1), 0), reinterpret_tensor(primals_11, (16, 4), (1, 16), 0), out=buf18)
buf19 = reinterpret_tensor(buf18, (4, 4, 4), (16, 4, 1), 0); del buf18 # reuse
# Topologically Sorted Source Nodes: [x_1, x_3, x_8], Original ATen: [aten.add]
triton_poi_fused_add_11.run(buf19, primals_3, buf12, primals_6, primals_12, 64, grid=grid(64), stream=stream0)
del primals_12
return (buf19, primals_3, primals_6, primals_7, reinterpret_tensor(buf2, (16, 4), (4, 1), 0), buf8, reinterpret_tensor(buf11, (16, 4), (4, 1), 0), buf12, reinterpret_tensor(buf15, (16, 4), (4, 1), 0), buf16, reinterpret_tensor(buf17, (16, 16), (16, 1), 0), primals_11, primals_9, primals_5, reinterpret_tensor(buf9, (16, 1, 4), (4, 1, 1), 0), reinterpret_tensor(buf4, (16, 1, 4), (4, 1, 1), 0), reinterpret_tensor(buf5, (16, 4, 1), (4, 1, 4), 0), primals_4, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((12, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_8 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_9 = rand_strided((16, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_10 = rand_strided((16, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_11 = rand_strided((4, 16), (16, 1), device='cuda:0', dtype=torch.float32)
primals_12 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9, primals_10, primals_11, primals_12])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch._C
import torch.serialization
from torch import nn
import torch.nn.functional as F
class DropPath(nn.Module):
"""Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks).
"""
def __init__(self, drop_prob=None):
super(DropPath, self).__init__()
self.drop_prob = drop_prob
def forward(self, x):
return drop_path(x, self.drop_prob, self.training)
def extra_repr(self) ->str:
return 'p={}'.format(self.drop_prob)
class Mlp(nn.Module):
def __init__(self, in_features, hidden_features=None, out_features=None,
act_layer=nn.GELU, drop=0.0):
super().__init__()
out_features = out_features or in_features
hidden_features = hidden_features or in_features
self.fc1 = nn.Linear(in_features, hidden_features)
self.act = act_layer()
self.fc2 = nn.Linear(hidden_features, out_features)
self.drop = nn.Dropout(drop)
def forward(self, x):
x = self.fc1(x)
x = self.act(x)
x = self.fc2(x)
x = self.drop(x)
return x
class Attention(nn.Module):
def __init__(self, dim, num_heads=8, qkv_bias=False, qk_scale=None,
attn_drop=0.0, proj_drop=0.0, window_size=None, attn_head_dim=None):
super().__init__()
self.num_heads = num_heads
head_dim = dim // num_heads
if attn_head_dim is not None:
head_dim = attn_head_dim
all_head_dim = head_dim * self.num_heads
self.scale = qk_scale or head_dim ** -0.5
self.qkv = nn.Linear(dim, all_head_dim * 3, bias=False)
if qkv_bias:
self.q_bias = nn.Parameter(torch.zeros(all_head_dim))
self.v_bias = nn.Parameter(torch.zeros(all_head_dim))
else:
self.q_bias = None
self.v_bias = None
if window_size:
self.window_size = window_size
self.num_relative_distance = (2 * window_size[0] - 1) * (2 *
window_size[1] - 1) + 3
self.relative_position_bias_table = nn.Parameter(torch.zeros(
self.num_relative_distance, num_heads))
coords_h = torch.arange(window_size[0])
coords_w = torch.arange(window_size[1])
coords = torch.stack(torch.meshgrid([coords_h, coords_w]))
coords_flatten = torch.flatten(coords, 1)
relative_coords = coords_flatten[:, :, None] - coords_flatten[:,
None, :]
relative_coords = relative_coords.permute(1, 2, 0).contiguous()
relative_coords[:, :, 0] += window_size[0] - 1
relative_coords[:, :, 1] += window_size[1] - 1
relative_coords[:, :, 0] *= 2 * window_size[1] - 1
relative_position_index = torch.zeros(size=(window_size[0] *
window_size[1] + 1,) * 2, dtype=relative_coords.dtype)
relative_position_index[1:, 1:] = relative_coords.sum(-1)
relative_position_index[0, 0:] = self.num_relative_distance - 3
relative_position_index[0:, 0] = self.num_relative_distance - 2
relative_position_index[0, 0] = self.num_relative_distance - 1
self.register_buffer('relative_position_index',
relative_position_index)
else:
self.window_size = None
self.relative_position_bias_table = None
self.relative_position_index = None
self.attn_drop = nn.Dropout(attn_drop)
self.proj = nn.Linear(all_head_dim, dim)
self.proj_drop = nn.Dropout(proj_drop)
def forward(self, x, rel_pos_bias=None):
B, N, _C = x.shape
qkv_bias = None
if self.q_bias is not None:
qkv_bias = torch.cat((self.q_bias, torch.zeros_like(self.v_bias,
requires_grad=False), self.v_bias))
qkv = F.linear(input=x, weight=self.qkv.weight, bias=qkv_bias)
qkv = qkv.reshape(B, N, 3, self.num_heads, -1).permute(2, 0, 3, 1, 4)
q, k, v = qkv[0], qkv[1], qkv[2]
q = q * self.scale
attn = q @ k.transpose(-2, -1)
if self.relative_position_bias_table is not None:
relative_position_bias = self.relative_position_bias_table[self
.relative_position_index.view(-1)].view(self.window_size[0] *
self.window_size[1] + 1, self.window_size[0] * self.
window_size[1] + 1, -1)
relative_position_bias = relative_position_bias.permute(2, 0, 1
).contiguous()
attn = attn + relative_position_bias.unsqueeze(0)
if rel_pos_bias is not None:
attn = attn + rel_pos_bias
attn = attn.softmax(dim=-1)
attn = self.attn_drop(attn)
x = (attn @ v).transpose(1, 2).reshape(B, N, -1)
x = self.proj(x)
x = self.proj_drop(x)
return x
class Block(nn.Module):
def __init__(self, dim, num_heads, mlp_ratio=4.0, qkv_bias=False,
qk_scale=None, drop=0.0, attn_drop=0.0, drop_path=0.0, init_values=
None, act_layer=nn.GELU, norm_layer=nn.LayerNorm, window_size=None,
attn_head_dim=None):
super().__init__()
self.norm1 = norm_layer(dim)
self.attn = Attention(dim, num_heads=num_heads, qkv_bias=qkv_bias,
qk_scale=qk_scale, attn_drop=attn_drop, proj_drop=drop,
window_size=window_size, attn_head_dim=attn_head_dim)
self.drop_path = DropPath(drop_path
) if drop_path > 0.0 else nn.Identity()
self.norm2 = norm_layer(dim)
mlp_hidden_dim = int(dim * mlp_ratio)
self.mlp = Mlp(in_features=dim, hidden_features=mlp_hidden_dim,
act_layer=act_layer, drop=drop)
if init_values is not None:
self.gamma_1 = nn.Parameter(init_values * torch.ones(dim),
requires_grad=True)
self.gamma_2 = nn.Parameter(init_values * torch.ones(dim),
requires_grad=True)
else:
self.gamma_1, self.gamma_2 = None, None
def forward(self, x, rel_pos_bias=None):
if self.gamma_1 is None:
x = x + self.drop_path(self.attn(self.norm1(x), rel_pos_bias=
rel_pos_bias))
x = x + self.drop_path(self.mlp(self.norm2(x)))
else:
x = x + self.drop_path(self.gamma_1 * self.attn(self.norm1(x),
rel_pos_bias=rel_pos_bias))
x = x + self.drop_path(self.gamma_2 * self.mlp(self.norm2(x)))
return x
def get_inputs():
return [torch.rand([4, 4, 4])]
def get_init_inputs():
return [[], {'dim': 4, 'num_heads': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
import torch._C
import torch.serialization
from torch import nn
import torch.nn.functional as F
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_native_layer_norm_0(in_ptr0, out_ptr0, out_ptr1,
xnumel, XBLOCK: tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + 4 * x0, xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr0 + (1 + 4 * x0), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr0 + (2 + 4 * x0), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (3 + 4 * x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp4 = tmp2 + tmp3
tmp6 = tmp4 + tmp5
tmp7 = 4.0
tmp8 = tmp6 / tmp7
tmp9 = tmp0 - tmp8
tmp10 = tmp9 * tmp9
tmp11 = tmp1 - tmp8
tmp12 = tmp11 * tmp11
tmp13 = tmp10 + tmp12
tmp14 = tmp3 - tmp8
tmp15 = tmp14 * tmp14
tmp16 = tmp13 + tmp15
tmp17 = tmp5 - tmp8
tmp18 = tmp17 * tmp17
tmp19 = tmp16 + tmp18
tmp20 = tmp19 / tmp7
tmp21 = 1e-05
tmp22 = tmp20 + tmp21
tmp23 = libdevice.rsqrt(tmp22)
tl.store(out_ptr0 + x0, tmp8, xmask)
tl.store(out_ptr1 + x0, tmp23, xmask)
@triton.jit
def triton_poi_fused_native_layer_norm_1(in_ptr0, in_ptr1, in_ptr2, in_ptr3,
in_ptr4, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
x0 = xindex % 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr1 + x1, xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr2 + x1, xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr3 + x0, xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr4 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 - tmp1
tmp4 = tmp2 * tmp3
tmp6 = tmp4 * tmp5
tmp8 = tmp6 + tmp7
tl.store(out_ptr0 + x2, tmp8, xmask)
@triton.jit
def triton_poi_fused_clone_mul_2(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK:
tl.constexpr, XBLOCK: tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = yindex // 4
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + 12 * x2 + 48 * y1), xmask & ymask,
eviction_policy='evict_last')
tmp1 = 1.0
tmp2 = tmp0 * tmp1
tl.store(out_ptr0 + (x2 + 4 * y3), tmp2, xmask & ymask)
@triton.jit
def triton_poi_fused_clone_3(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK: tl.
constexpr, XBLOCK: tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = yindex // 4
y3 = yindex
tmp0 = tl.load(in_ptr0 + (4 + y0 + 12 * x2 + 48 * y1), xmask & ymask,
eviction_policy='evict_last')
tl.store(out_ptr0 + (x2 + 4 * y3), tmp0, xmask & ymask)
@triton.jit
def triton_poi_fused__softmax_4(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp3 = triton_helpers.maximum(tmp1, tmp2)
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp7 = triton_helpers.maximum(tmp5, tmp6)
tmp8 = tmp0 - tmp7
tmp9 = tl_math.exp(tmp8)
tl.store(out_ptr0 + x2, tmp9, xmask)
@triton.jit
def triton_poi_fused__softmax_5(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + x2, tmp8, xmask)
@triton.jit
def triton_poi_fused_clone_6(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK: tl.
constexpr, XBLOCK: tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = yindex // 4
y3 = yindex
tmp0 = tl.load(in_ptr0 + (8 + y0 + 12 * x2 + 48 * y1), xmask & ymask,
eviction_policy='evict_last')
tl.store(out_ptr0 + (x2 + 4 * y3), tmp0, xmask & ymask)
@triton.jit
def triton_poi_fused_clone_7(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK: tl.
constexpr, XBLOCK: tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = yindex // 4
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + 4 * x2 + 16 * y1), xmask & ymask,
eviction_policy='evict_last')
tl.store(out_ptr0 + (x2 + 4 * y3), tmp0, xmask & ymask)
@triton.jit
def triton_poi_fused_add_native_layer_norm_8(in_ptr0, in_ptr1, in_ptr2,
out_ptr0, out_ptr1, xnumel, XBLOCK: tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + 4 * x0, xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + 4 * x0, xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr2 + 0)
tmp3 = tl.broadcast_to(tmp2, [XBLOCK])
tmp6 = tl.load(in_ptr0 + (1 + 4 * x0), xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr1 + (1 + 4 * x0), xmask, eviction_policy='evict_last')
tmp8 = tl.load(in_ptr2 + 1)
tmp9 = tl.broadcast_to(tmp8, [XBLOCK])
tmp13 = tl.load(in_ptr0 + (2 + 4 * x0), xmask, eviction_policy='evict_last'
)
tmp14 = tl.load(in_ptr1 + (2 + 4 * x0), xmask, eviction_policy='evict_last'
)
tmp15 = tl.load(in_ptr2 + 2)
tmp16 = tl.broadcast_to(tmp15, [XBLOCK])
tmp20 = tl.load(in_ptr0 + (3 + 4 * x0), xmask, eviction_policy='evict_last'
)
tmp21 = tl.load(in_ptr1 + (3 + 4 * x0), xmask, eviction_policy='evict_last'
)
tmp22 = tl.load(in_ptr2 + 3)
tmp23 = tl.broadcast_to(tmp22, [XBLOCK])
tmp4 = tmp1 + tmp3
tmp5 = tmp0 + tmp4
tmp10 = tmp7 + tmp9
tmp11 = tmp6 + tmp10
tmp12 = tmp5 + tmp11
tmp17 = tmp14 + tmp16
tmp18 = tmp13 + tmp17
tmp19 = tmp12 + tmp18
tmp24 = tmp21 + tmp23
tmp25 = tmp20 + tmp24
tmp26 = tmp19 + tmp25
tmp27 = 4.0
tmp28 = tmp26 / tmp27
tmp29 = tmp5 - tmp28
tmp30 = tmp29 * tmp29
tmp31 = tmp11 - tmp28
tmp32 = tmp31 * tmp31
tmp33 = tmp30 + tmp32
tmp34 = tmp18 - tmp28
tmp35 = tmp34 * tmp34
tmp36 = tmp33 + tmp35
tmp37 = tmp25 - tmp28
tmp38 = tmp37 * tmp37
tmp39 = tmp36 + tmp38
tmp40 = tmp39 / tmp27
tl.store(out_ptr0 + x0, tmp28, xmask)
tl.store(out_ptr1 + x0, tmp40, xmask)
@triton.jit
def triton_poi_fused_add_native_layer_norm_9(in_ptr0, in_ptr1, in_ptr2,
in_ptr3, in_ptr4, in_ptr5, in_ptr6, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr1 + x2, xmask)
tmp2 = tl.load(in_ptr2 + x0, xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr3 + x1, xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr4 + x1, xmask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr5 + x0, xmask, eviction_policy='evict_last')
tmp14 = tl.load(in_ptr6 + x0, xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp4 = tmp0 + tmp3
tmp6 = tmp4 - tmp5
tmp8 = 1e-05
tmp9 = tmp7 + tmp8
tmp10 = libdevice.rsqrt(tmp9)
tmp11 = tmp6 * tmp10
tmp13 = tmp11 * tmp12
tmp15 = tmp13 + tmp14
tl.store(out_ptr0 + x2, tmp15, xmask)
@triton.jit
def triton_poi_fused_gelu_10(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + x0, xmask)
tmp1 = 0.5
tmp2 = tmp0 * tmp1
tmp3 = 0.7071067811865476
tmp4 = tmp0 * tmp3
tmp5 = libdevice.erf(tmp4)
tmp6 = 1.0
tmp7 = tmp5 + tmp6
tmp8 = tmp2 * tmp7
tl.store(out_ptr0 + x0, tmp8, xmask)
@triton.jit
def triton_poi_fused_add_11(in_out_ptr0, in_ptr0, in_ptr1, in_ptr2, in_ptr3,
xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr1 + x2, xmask)
tmp2 = tl.load(in_ptr2 + x0, xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_out_ptr0 + x2, xmask)
tmp6 = tl.load(in_ptr3 + x0, xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp4 = tmp0 + tmp3
tmp7 = tmp5 + tmp6
tmp8 = tmp4 + tmp7
tl.store(in_out_ptr0 + x2, tmp8, xmask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7, primals_8, primals_9, primals_10, primals_11, primals_12
) = args
args.clear()
assert_size_stride(primals_1, (4,), (1,))
assert_size_stride(primals_2, (4,), (1,))
assert_size_stride(primals_3, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_4, (12, 4), (4, 1))
assert_size_stride(primals_5, (4, 4), (4, 1))
assert_size_stride(primals_6, (4,), (1,))
assert_size_stride(primals_7, (4,), (1,))
assert_size_stride(primals_8, (4,), (1,))
assert_size_stride(primals_9, (16, 4), (4, 1))
assert_size_stride(primals_10, (16,), (1,))
assert_size_stride(primals_11, (4, 16), (16, 1))
assert_size_stride(primals_12, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 1), (4, 1, 16), torch.float32)
buf1 = empty_strided_cuda((4, 4, 1), (4, 1, 16), torch.float32)
get_raw_stream(0)
triton_poi_fused_native_layer_norm_0[grid(16)](primals_3, buf0,
buf1, 16, XBLOCK=16, num_warps=1, num_stages=1)
buf2 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
triton_poi_fused_native_layer_norm_1[grid(64)](primals_3, buf0,
buf1, primals_1, primals_2, buf2, 64, XBLOCK=64, num_warps=1,
num_stages=1)
del primals_1
del primals_2
buf3 = empty_strided_cuda((16, 12), (12, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(buf2, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_4, (4, 12), (1, 4), 0), out=buf3)
buf4 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 1), torch.float32)
triton_poi_fused_clone_mul_2[grid(16, 4)](buf3, buf4, 16, 4, XBLOCK
=4, YBLOCK=16, num_warps=1, num_stages=1)
buf5 = empty_strided_cuda((4, 4, 1, 4), (16, 4, 4, 1), torch.float32)
triton_poi_fused_clone_3[grid(16, 4)](buf3, buf5, 16, 4, XBLOCK=4,
YBLOCK=16, num_warps=1, num_stages=1)
buf6 = empty_strided_cuda((16, 4, 4), (16, 4, 1), torch.float32)
extern_kernels.bmm(reinterpret_tensor(buf4, (16, 4, 1), (4, 1, 0),
0), reinterpret_tensor(buf5, (16, 1, 4), (4, 0, 1), 0), out=buf6)
buf7 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
triton_poi_fused__softmax_4[grid(256)](buf6, buf7, 256, XBLOCK=128,
num_warps=4, num_stages=1)
buf8 = reinterpret_tensor(buf6, (4, 4, 4, 4), (64, 16, 4, 1), 0)
del buf6
triton_poi_fused__softmax_5[grid(256)](buf7, buf8, 256, XBLOCK=256,
num_warps=4, num_stages=1)
buf9 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 1), torch.float32)
triton_poi_fused_clone_6[grid(16, 4)](buf3, buf9, 16, 4, XBLOCK=4,
YBLOCK=16, num_warps=1, num_stages=1)
del buf3
buf10 = empty_strided_cuda((16, 4, 1), (4, 1, 1), torch.float32)
extern_kernels.bmm(reinterpret_tensor(buf8, (16, 4, 4), (16, 4, 1),
0), reinterpret_tensor(buf9, (16, 4, 1), (4, 1, 0), 0), out=buf10)
buf11 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
triton_poi_fused_clone_7[grid(16, 4)](buf10, buf11, 16, 4, XBLOCK=4,
YBLOCK=16, num_warps=1, num_stages=1)
buf12 = reinterpret_tensor(buf10, (16, 4), (4, 1), 0)
del buf10
extern_kernels.mm(reinterpret_tensor(buf11, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_5, (4, 4), (1, 4), 0), out=buf12)
buf13 = buf1
del buf1
buf14 = buf0
del buf0
triton_poi_fused_add_native_layer_norm_8[grid(16)](primals_3, buf12,
primals_6, buf13, buf14, 16, XBLOCK=16, num_warps=1, num_stages=1)
buf15 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
triton_poi_fused_add_native_layer_norm_9[grid(64)](primals_3, buf12,
primals_6, buf13, buf14, primals_7, primals_8, buf15, 64,
XBLOCK=64, num_warps=1, num_stages=1)
del buf13
del buf14
del primals_8
buf16 = reinterpret_tensor(buf7, (16, 16), (16, 1), 0)
del buf7
extern_kernels.addmm(primals_10, reinterpret_tensor(buf15, (16, 4),
(4, 1), 0), reinterpret_tensor(primals_9, (4, 16), (1, 4), 0),
alpha=1, beta=1, out=buf16)
del primals_10
buf17 = empty_strided_cuda((4, 4, 16), (64, 16, 1), torch.float32)
triton_poi_fused_gelu_10[grid(256)](buf16, buf17, 256, XBLOCK=128,
num_warps=4, num_stages=1)
buf18 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(buf17, (16, 16), (16, 1), 0),
reinterpret_tensor(primals_11, (16, 4), (1, 16), 0), out=buf18)
buf19 = reinterpret_tensor(buf18, (4, 4, 4), (16, 4, 1), 0)
del buf18
triton_poi_fused_add_11[grid(64)](buf19, primals_3, buf12,
primals_6, primals_12, 64, XBLOCK=64, num_warps=1, num_stages=1)
del primals_12
return buf19, primals_3, primals_6, primals_7, reinterpret_tensor(buf2,
(16, 4), (4, 1), 0), buf8, reinterpret_tensor(buf11, (16, 4), (4, 1), 0
), buf12, reinterpret_tensor(buf15, (16, 4), (4, 1), 0
), buf16, reinterpret_tensor(buf17, (16, 16), (16, 1), 0
), primals_11, primals_9, primals_5, reinterpret_tensor(buf9, (16,
1, 4), (4, 1, 1), 0), reinterpret_tensor(buf4, (16, 1, 4), (4, 1, 1), 0
), reinterpret_tensor(buf5, (16, 4, 1), (4, 1, 4), 0), primals_4
class DropPath(nn.Module):
"""Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks).
"""
def __init__(self, drop_prob=None):
super(DropPath, self).__init__()
self.drop_prob = drop_prob
def forward(self, x):
return drop_path(x, self.drop_prob, self.training)
def extra_repr(self) ->str:
return 'p={}'.format(self.drop_prob)
class Mlp(nn.Module):
def __init__(self, in_features, hidden_features=None, out_features=None,
act_layer=nn.GELU, drop=0.0):
super().__init__()
out_features = out_features or in_features
hidden_features = hidden_features or in_features
self.fc1 = nn.Linear(in_features, hidden_features)
self.act = act_layer()
self.fc2 = nn.Linear(hidden_features, out_features)
self.drop = nn.Dropout(drop)
def forward(self, x):
x = self.fc1(x)
x = self.act(x)
x = self.fc2(x)
x = self.drop(x)
return x
class Attention(nn.Module):
def __init__(self, dim, num_heads=8, qkv_bias=False, qk_scale=None,
attn_drop=0.0, proj_drop=0.0, window_size=None, attn_head_dim=None):
super().__init__()
self.num_heads = num_heads
head_dim = dim // num_heads
if attn_head_dim is not None:
head_dim = attn_head_dim
all_head_dim = head_dim * self.num_heads
self.scale = qk_scale or head_dim ** -0.5
self.qkv = nn.Linear(dim, all_head_dim * 3, bias=False)
if qkv_bias:
self.q_bias = nn.Parameter(torch.zeros(all_head_dim))
self.v_bias = nn.Parameter(torch.zeros(all_head_dim))
else:
self.q_bias = None
self.v_bias = None
if window_size:
self.window_size = window_size
self.num_relative_distance = (2 * window_size[0] - 1) * (2 *
window_size[1] - 1) + 3
self.relative_position_bias_table = nn.Parameter(torch.zeros(
self.num_relative_distance, num_heads))
coords_h = torch.arange(window_size[0])
coords_w = torch.arange(window_size[1])
coords = torch.stack(torch.meshgrid([coords_h, coords_w]))
coords_flatten = torch.flatten(coords, 1)
relative_coords = coords_flatten[:, :, None] - coords_flatten[:,
None, :]
relative_coords = relative_coords.permute(1, 2, 0).contiguous()
relative_coords[:, :, 0] += window_size[0] - 1
relative_coords[:, :, 1] += window_size[1] - 1
relative_coords[:, :, 0] *= 2 * window_size[1] - 1
relative_position_index = torch.zeros(size=(window_size[0] *
window_size[1] + 1,) * 2, dtype=relative_coords.dtype)
relative_position_index[1:, 1:] = relative_coords.sum(-1)
relative_position_index[0, 0:] = self.num_relative_distance - 3
relative_position_index[0:, 0] = self.num_relative_distance - 2
relative_position_index[0, 0] = self.num_relative_distance - 1
self.register_buffer('relative_position_index',
relative_position_index)
else:
self.window_size = None
self.relative_position_bias_table = None
self.relative_position_index = None
self.attn_drop = nn.Dropout(attn_drop)
self.proj = nn.Linear(all_head_dim, dim)
self.proj_drop = nn.Dropout(proj_drop)
def forward(self, x, rel_pos_bias=None):
B, N, _C = x.shape
qkv_bias = None
if self.q_bias is not None:
qkv_bias = torch.cat((self.q_bias, torch.zeros_like(self.v_bias,
requires_grad=False), self.v_bias))
qkv = F.linear(input=x, weight=self.qkv.weight, bias=qkv_bias)
qkv = qkv.reshape(B, N, 3, self.num_heads, -1).permute(2, 0, 3, 1, 4)
q, k, v = qkv[0], qkv[1], qkv[2]
q = q * self.scale
attn = q @ k.transpose(-2, -1)
if self.relative_position_bias_table is not None:
relative_position_bias = self.relative_position_bias_table[self
.relative_position_index.view(-1)].view(self.window_size[0] *
self.window_size[1] + 1, self.window_size[0] * self.
window_size[1] + 1, -1)
relative_position_bias = relative_position_bias.permute(2, 0, 1
).contiguous()
attn = attn + relative_position_bias.unsqueeze(0)
if rel_pos_bias is not None:
attn = attn + rel_pos_bias
attn = attn.softmax(dim=-1)
attn = self.attn_drop(attn)
x = (attn @ v).transpose(1, 2).reshape(B, N, -1)
x = self.proj(x)
x = self.proj_drop(x)
return x
class BlockNew(nn.Module):
def __init__(self, dim, num_heads, mlp_ratio=4.0, qkv_bias=False,
qk_scale=None, drop=0.0, attn_drop=0.0, drop_path=0.0, init_values=
None, act_layer=nn.GELU, norm_layer=nn.LayerNorm, window_size=None,
attn_head_dim=None):
super().__init__()
self.norm1 = norm_layer(dim)
self.attn = Attention(dim, num_heads=num_heads, qkv_bias=qkv_bias,
qk_scale=qk_scale, attn_drop=attn_drop, proj_drop=drop,
window_size=window_size, attn_head_dim=attn_head_dim)
self.drop_path = DropPath(drop_path
) if drop_path > 0.0 else nn.Identity()
self.norm2 = norm_layer(dim)
mlp_hidden_dim = int(dim * mlp_ratio)
self.mlp = Mlp(in_features=dim, hidden_features=mlp_hidden_dim,
act_layer=act_layer, drop=drop)
if init_values is not None:
self.gamma_1 = nn.Parameter(init_values * torch.ones(dim),
requires_grad=True)
self.gamma_2 = nn.Parameter(init_values * torch.ones(dim),
requires_grad=True)
else:
self.gamma_1, self.gamma_2 = None, None
def forward(self, input_0):
primals_1 = self.norm1.weight
primals_2 = self.norm1.bias
primals_4 = self.attn.qkv.weight
primals_5 = self.attn.proj.weight
primals_6 = self.attn.proj.bias
primals_7 = self.norm2.weight
primals_8 = self.norm2.bias
primals_9 = self.mlp.fc1.weight
primals_10 = self.mlp.fc1.bias
primals_11 = self.mlp.fc2.weight
primals_12 = self.mlp.fc2.bias
primals_3 = input_0
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7, primals_8, primals_9,
primals_10, primals_11, primals_12])
return output[0]
| huazai-1994/24th-resolution-for-STAC-Overflow | Block | false | 6,848 | [
"Apache-2.0"
] | 1 | 80bb3b367a126264823ffc597dc01586c262f9d9 | https://github.com/huazai-1994/24th-resolution-for-STAC-Overflow/tree/80bb3b367a126264823ffc597dc01586c262f9d9 | import torch
import torch._C
import torch.serialization
from torch import nn
import torch.nn.functional as F
class DropPath(nn.Module):
"""Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks).
"""
def __init__(self, drop_prob=None):
super().__init__()
self.drop_prob = drop_prob
def forward(self, x):
return drop_path(x, self.drop_prob, self.training)
def extra_repr(self) ->str:
return 'p={}'.format(self.drop_prob)
class Mlp(nn.Module):
def __init__(self, in_features, hidden_features=None, out_features=None,
act_layer=nn.GELU, drop=0.0):
super().__init__()
out_features = out_features or in_features
hidden_features = hidden_features or in_features
self.fc1 = nn.Linear(in_features, hidden_features)
self.act = act_layer()
self.fc2 = nn.Linear(hidden_features, out_features)
self.drop = nn.Dropout(drop)
def forward(self, x):
x = self.fc1(x)
x = self.act(x)
x = self.fc2(x)
x = self.drop(x)
return x
class Attention(nn.Module):
def __init__(self, dim, num_heads=8, qkv_bias=False, qk_scale=None,
attn_drop=0.0, proj_drop=0.0, window_size=None, attn_head_dim=None):
super().__init__()
self.num_heads = num_heads
head_dim = dim // num_heads
if attn_head_dim is not None:
head_dim = attn_head_dim
all_head_dim = head_dim * self.num_heads
self.scale = qk_scale or head_dim ** -0.5
self.qkv = nn.Linear(dim, all_head_dim * 3, bias=False)
if qkv_bias:
self.q_bias = nn.Parameter(torch.zeros(all_head_dim))
self.v_bias = nn.Parameter(torch.zeros(all_head_dim))
else:
self.q_bias = None
self.v_bias = None
if window_size:
self.window_size = window_size
self.num_relative_distance = (2 * window_size[0] - 1) * (2 *
window_size[1] - 1) + 3
self.relative_position_bias_table = nn.Parameter(torch.zeros(
self.num_relative_distance, num_heads))
coords_h = torch.arange(window_size[0])
coords_w = torch.arange(window_size[1])
coords = torch.stack(torch.meshgrid([coords_h, coords_w]))
coords_flatten = torch.flatten(coords, 1)
relative_coords = coords_flatten[:, :, None] - coords_flatten[:,
None, :]
relative_coords = relative_coords.permute(1, 2, 0).contiguous()
relative_coords[:, :, 0] += window_size[0] - 1
relative_coords[:, :, 1] += window_size[1] - 1
relative_coords[:, :, 0] *= 2 * window_size[1] - 1
relative_position_index = torch.zeros(size=(window_size[0] *
window_size[1] + 1,) * 2, dtype=relative_coords.dtype)
relative_position_index[1:, 1:] = relative_coords.sum(-1)
relative_position_index[0, 0:] = self.num_relative_distance - 3
relative_position_index[0:, 0] = self.num_relative_distance - 2
relative_position_index[0, 0] = self.num_relative_distance - 1
self.register_buffer('relative_position_index',
relative_position_index)
else:
self.window_size = None
self.relative_position_bias_table = None
self.relative_position_index = None
self.attn_drop = nn.Dropout(attn_drop)
self.proj = nn.Linear(all_head_dim, dim)
self.proj_drop = nn.Dropout(proj_drop)
def forward(self, x, rel_pos_bias=None):
B, N, _C = x.shape
qkv_bias = None
if self.q_bias is not None:
qkv_bias = torch.cat((self.q_bias, torch.zeros_like(self.v_bias,
requires_grad=False), self.v_bias))
qkv = F.linear(input=x, weight=self.qkv.weight, bias=qkv_bias)
qkv = qkv.reshape(B, N, 3, self.num_heads, -1).permute(2, 0, 3, 1,
# ... truncated (>4000 chars) for memory efficiency |
EncoderBlock | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/xx/cxxsy7adcakcv5c6imnaqsvf3ebhgbph77vaoembzakbqdrjycbc.py
# Topologically Sorted Source Nodes: [mean, std, sub, mul, add, truediv, input_norm], Original ATen: [aten.mean, aten.std, aten.sub, aten.mul, aten.add, aten.div]
# Source node to ATen node mapping:
# add => add
# input_norm => add_1
# mean => mean
# mul => mul
# std => sqrt, var
# sub => sub
# truediv => div
# Graph fragment:
# %mean : [num_users=1] = call_function[target=torch.ops.aten.mean.dim](args = (%primals_1, [-1], True), kwargs = {})
# %var : [num_users=1] = call_function[target=torch.ops.aten.var.correction](args = (%primals_1, [-1]), kwargs = {correction: 1.0, keepdim: True})
# %sqrt : [num_users=1] = call_function[target=torch.ops.aten.sqrt.default](args = (%var,), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%primals_1, %mean), kwargs = {})
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%primals_2, %sub), kwargs = {})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%sqrt, 1e-06), kwargs = {})
# %div : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%mul, %add), kwargs = {})
# %add_1 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%div, %primals_3), kwargs = {})
triton_poi_fused_add_div_mean_mul_std_sub_0 = async_compile.triton('triton_poi_fused_add_div_mean_mul_std_sub_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_div_mean_mul_std_sub_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 7, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_div_mean_mul_std_sub_0(in_ptr0, in_ptr1, in_ptr2, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 4
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (x2), xmask)
tmp2 = tl.load(in_ptr1 + (4*x1), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr1 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr1 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr1 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp30 = tl.load(in_ptr2 + (x0), xmask, eviction_policy='evict_last')
tmp4 = tmp2 + tmp3
tmp6 = tmp4 + tmp5
tmp8 = tmp6 + tmp7
tmp9 = 4.0
tmp10 = tmp8 / tmp9
tmp11 = tmp1 - tmp10
tmp12 = tmp0 * tmp11
tmp13 = tmp2 - tmp10
tmp14 = tmp13 * tmp13
tmp15 = tmp3 - tmp10
tmp16 = tmp15 * tmp15
tmp17 = tmp14 + tmp16
tmp18 = tmp5 - tmp10
tmp19 = tmp18 * tmp18
tmp20 = tmp17 + tmp19
tmp21 = tmp7 - tmp10
tmp22 = tmp21 * tmp21
tmp23 = tmp20 + tmp22
tmp24 = 3.0
tmp25 = tmp23 / tmp24
tmp26 = libdevice.sqrt(tmp25)
tmp27 = 1e-06
tmp28 = tmp26 + tmp27
tmp29 = tmp12 / tmp28
tmp31 = tmp29 + tmp30
tl.store(out_ptr0 + (x2), tmp31, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/wc/cwc2g4aer7dhpqcbpvbocvjyfzzfliu4x7goxlv7aksmsfu7j6nh.py
# Topologically Sorted Source Nodes: [query_up_1, scores], Original ATen: [aten.div, aten.clone]
# Source node to ATen node mapping:
# query_up_1 => div_1
# scores => clone
# Graph fragment:
# %div_1 : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%permute_5, 1.0), kwargs = {})
# %clone : [num_users=1] = call_function[target=torch.ops.aten.clone.default](args = (%expand,), kwargs = {memory_format: torch.contiguous_format})
triton_poi_fused_clone_div_1 = async_compile.triton('triton_poi_fused_clone_div_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16, 4], tile_hint=TileHint.DEFAULT,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clone_div_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clone_div_1(in_ptr0, in_ptr1, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = (yindex // 4)
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + (4*x2) + (16*y1)), xmask & ymask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (y0), ymask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = 1.0
tmp4 = tmp2 * tmp3
tl.store(out_ptr0 + (x2 + (4*y3)), tmp4, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/7k/c7kkxqo5r65gqykuvge3exgf3trgxmm4raf7gypitw4ynuylbeao.py
# Topologically Sorted Source Nodes: [scores], Original ATen: [aten.clone]
# Source node to ATen node mapping:
# scores => clone_1
# Graph fragment:
# %clone_1 : [num_users=1] = call_function[target=torch.ops.aten.clone.default](args = (%expand_1,), kwargs = {memory_format: torch.contiguous_format})
triton_poi_fused_clone_2 = async_compile.triton('triton_poi_fused_clone_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16, 4], tile_hint=TileHint.DEFAULT,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clone_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clone_2(in_ptr0, in_ptr1, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = (yindex // 4)
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + (4*x2) + (16*y1)), xmask & ymask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (y0), ymask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(out_ptr0 + (x2 + (4*y3)), tmp2, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/s4/cs472yivvl3yzse325afzknsz7ua5dqrqzmwls3lwujk3hte6xkl.py
# Topologically Sorted Source Nodes: [attn], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# attn => amax, exp, sub_1
# Graph fragment:
# %amax : [num_users=1] = call_function[target=torch.ops.aten.amax.default](args = (%view_11, [-1], True), kwargs = {})
# %sub_1 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%view_11, %amax), kwargs = {})
# %exp : [num_users=2] = call_function[target=torch.ops.aten.exp.default](args = (%sub_1,), kwargs = {})
triton_poi_fused__softmax_3 = async_compile.triton('triton_poi_fused__softmax_3', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_3', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_3(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp3 = triton_helpers.maximum(tmp1, tmp2)
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp7 = triton_helpers.maximum(tmp5, tmp6)
tmp8 = tmp0 - tmp7
tmp9 = tl_math.exp(tmp8)
tl.store(out_ptr0 + (x2), tmp9, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/zh/czh6tw7ngffcygnivwvcjex5edxy3ms4t27ymyn2hemxlpspxzq7.py
# Topologically Sorted Source Nodes: [attn], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# attn => div_2, sum_1
# Graph fragment:
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%exp, [-1], True), kwargs = {})
# %div_2 : [num_users=2] = call_function[target=torch.ops.aten.div.Tensor](args = (%exp, %sum_1), kwargs = {})
triton_poi_fused__softmax_4 = async_compile.triton('triton_poi_fused__softmax_4', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_4', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_4(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + (x2), tmp8, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/we/cwe54p4p4jvwbdktkpj3wy2coheu6f3r3dgvi7ozm7xjfk4mgbwx.py
# Topologically Sorted Source Nodes: [contiguous], Original ATen: [aten.clone]
# Source node to ATen node mapping:
# contiguous => clone_4
# Graph fragment:
# %clone_4 : [num_users=1] = call_function[target=torch.ops.aten.clone.default](args = (%permute_7,), kwargs = {memory_format: torch.contiguous_format})
triton_poi_fused_clone_5 = async_compile.triton('triton_poi_fused_clone_5', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16, 4], tile_hint=TileHint.SQUARE,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clone_5', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clone_5(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = (yindex // 4)
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + (4*x2) + (16*y1)), xmask & ymask, eviction_policy='evict_last')
tl.store(out_ptr0 + (x2 + (4*y3)), tmp0, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/k3/ck3fooheqiq5spsm2bt5dle5th5oggkmlao36xduvmqkrltazlaw.py
# Topologically Sorted Source Nodes: [out, mean_2, std_2], Original ATen: [aten.add, aten.mean, aten.std]
# Source node to ATen node mapping:
# mean_2 => mean_1
# out => add_2
# std_2 => var_1
# Graph fragment:
# %add_2 : [num_users=4] = call_function[target=torch.ops.aten.add.Tensor](args = (%view_17, %primals_1), kwargs = {})
# %mean_1 : [num_users=1] = call_function[target=torch.ops.aten.mean.dim](args = (%add_2, [-1], True), kwargs = {})
# %var_1 : [num_users=1] = call_function[target=torch.ops.aten.var.correction](args = (%add_2, [-1]), kwargs = {correction: 1.0, keepdim: True})
triton_poi_fused_add_mean_std_6 = async_compile.triton('triton_poi_fused_add_mean_std_6', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_mean_std_6', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 8, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_mean_std_6(in_out_ptr0, in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (4*x0), xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (4*x0), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr0 + (1 + (4*x0)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr1 + (1 + (4*x0)), xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr0 + (2 + (4*x0)), xmask, eviction_policy='evict_last')
tmp8 = tl.load(in_ptr1 + (2 + (4*x0)), xmask, eviction_policy='evict_last')
tmp11 = tl.load(in_ptr0 + (3 + (4*x0)), xmask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr1 + (3 + (4*x0)), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp5 = tmp3 + tmp4
tmp6 = tmp2 + tmp5
tmp9 = tmp7 + tmp8
tmp10 = tmp6 + tmp9
tmp13 = tmp11 + tmp12
tmp14 = tmp10 + tmp13
tmp15 = 4.0
tmp16 = tmp14 / tmp15
tmp17 = tmp2 - tmp16
tmp18 = tmp17 * tmp17
tmp19 = tmp5 - tmp16
tmp20 = tmp19 * tmp19
tmp21 = tmp18 + tmp20
tmp22 = tmp9 - tmp16
tmp23 = tmp22 * tmp22
tmp24 = tmp21 + tmp23
tmp25 = tmp13 - tmp16
tmp26 = tmp25 * tmp25
tmp27 = tmp24 + tmp26
tmp28 = 3.0
tmp29 = tmp27 / tmp28
tl.store(in_out_ptr0 + (x0), tmp29, xmask)
tl.store(out_ptr0 + (x0), tmp16, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/44/c44nsundpysjinc27mkdhx3tcd7nqj5ir3mp6sd7nqx3ryqium46.py
# Topologically Sorted Source Nodes: [out, mean_2, std_2, sub_1, mul_1, add_3, truediv_2, add_4], Original ATen: [aten.add, aten.mean, aten.std, aten.sub, aten.mul, aten.div]
# Source node to ATen node mapping:
# add_3 => add_3
# add_4 => add_4
# mean_2 => mean_1
# mul_1 => mul_1
# out => add_2
# std_2 => sqrt_1
# sub_1 => sub_2
# truediv_2 => div_3
# Graph fragment:
# %add_2 : [num_users=4] = call_function[target=torch.ops.aten.add.Tensor](args = (%view_17, %primals_1), kwargs = {})
# %mean_1 : [num_users=1] = call_function[target=torch.ops.aten.mean.dim](args = (%add_2, [-1], True), kwargs = {})
# %sqrt_1 : [num_users=1] = call_function[target=torch.ops.aten.sqrt.default](args = (%var_1,), kwargs = {})
# %sub_2 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%add_2, %mean_1), kwargs = {})
# %mul_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%primals_12, %sub_2), kwargs = {})
# %add_3 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%sqrt_1, 1e-06), kwargs = {})
# %div_3 : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%mul_1, %add_3), kwargs = {})
# %add_4 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%div_3, %primals_13), kwargs = {})
triton_poi_fused_add_div_mean_mul_std_sub_7 = async_compile.triton('triton_poi_fused_add_div_mean_mul_std_sub_7', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: '*fp32', 5: '*fp32', 6: '*fp32', 7: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5, 6, 7), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_div_mean_mul_std_sub_7', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 6, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_div_mean_mul_std_sub_7(in_ptr0, in_ptr1, in_ptr2, in_ptr3, in_ptr4, in_ptr5, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 4
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (x2), xmask)
tmp2 = tl.load(in_ptr2 + (x2), xmask)
tmp4 = tl.load(in_ptr3 + (x1), xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr4 + (x1), xmask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr5 + (x0), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 - tmp4
tmp6 = tmp0 * tmp5
tmp8 = libdevice.sqrt(tmp7)
tmp9 = 1e-06
tmp10 = tmp8 + tmp9
tmp11 = tmp6 / tmp10
tmp13 = tmp11 + tmp12
tl.store(out_ptr0 + (x2), tmp13, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/23/c23j2vk753qpctrm5kblwdo7f2zh4pnjylmlrcdhyl7syfjonudr.py
# Topologically Sorted Source Nodes: [relu], Original ATen: [aten.relu, aten.threshold_backward]
# Source node to ATen node mapping:
# relu => relu
# Graph fragment:
# %relu : [num_users=1] = call_function[target=torch.ops.aten.relu.default](args = (%view_20,), kwargs = {})
# %le : [num_users=1] = call_function[target=torch.ops.aten.le.Scalar](args = (%view_25, 0), kwargs = {})
triton_poi_fused_relu_threshold_backward_8 = async_compile.triton('triton_poi_fused_relu_threshold_backward_8', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*i1', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_relu_threshold_backward_8', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_relu_threshold_backward_8(in_out_ptr0, in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(in_out_ptr0 + (x2), tmp4, xmask)
tl.store(out_ptr0 + (x2), tmp6, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/gt/cgtudwfzrgd5etmagoppxnphqs7ogp52v7457gc6gno6nqbx34nx.py
# Topologically Sorted Source Nodes: [linear_5], Original ATen: [aten.view]
# Source node to ATen node mapping:
# linear_5 => view_26
# Graph fragment:
# %view_26 : [num_users=2] = call_function[target=torch.ops.aten.reshape.default](args = (%view_25, [16, 4]), kwargs = {})
triton_poi_fused_view_9 = async_compile.triton('triton_poi_fused_view_9', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_view_9', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_view_9(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 4
x1 = (xindex // 4)
x2 = xindex
tmp0 = tl.load(in_ptr0 + (x0 + (4*x1) + (16*((x1 % 4) // 4))), xmask)
tl.store(out_ptr0 + (x2), tmp0, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/jj/cjjjzfxy5xd234tvi74ytzgml2jbf2lzikobqebf5uzkducoyjmg.py
# Topologically Sorted Source Nodes: [out, add_5], Original ATen: [aten.add]
# Source node to ATen node mapping:
# add_5 => add_5
# out => add_2
# Graph fragment:
# %add_2 : [num_users=4] = call_function[target=torch.ops.aten.add.Tensor](args = (%view_17, %primals_1), kwargs = {})
# %add_5 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%view_27, %add_2), kwargs = {})
triton_poi_fused_add_10 = async_compile.triton('triton_poi_fused_add_10', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_10', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 4, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_10(in_out_ptr0, in_ptr0, in_ptr1, in_ptr2, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr1 + (x2), xmask)
tmp4 = tl.load(in_ptr2 + (x2), xmask)
tmp2 = tmp0 + tmp1
tmp5 = tmp3 + tmp4
tmp6 = tmp2 + tmp5
tl.store(in_out_ptr0 + (x2), tmp6, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9, primals_10, primals_11, primals_12, primals_13, primals_14, primals_15, primals_16, primals_17 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_2, (4, ), (1, ))
assert_size_stride(primals_3, (4, ), (1, ))
assert_size_stride(primals_4, (4, 4), (4, 1))
assert_size_stride(primals_5, (4, ), (1, ))
assert_size_stride(primals_6, (4, 4), (4, 1))
assert_size_stride(primals_7, (4, ), (1, ))
assert_size_stride(primals_8, (4, 4), (4, 1))
assert_size_stride(primals_9, (4, ), (1, ))
assert_size_stride(primals_10, (4, 4), (4, 1))
assert_size_stride(primals_11, (4, ), (1, ))
assert_size_stride(primals_12, (4, ), (1, ))
assert_size_stride(primals_13, (4, ), (1, ))
assert_size_stride(primals_14, (4, 4), (4, 1))
assert_size_stride(primals_15, (4, ), (1, ))
assert_size_stride(primals_16, (4, 4), (4, 1))
assert_size_stride(primals_17, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [mean, std, sub, mul, add, truediv, input_norm], Original ATen: [aten.mean, aten.std, aten.sub, aten.mul, aten.add, aten.div]
stream0 = get_raw_stream(0)
triton_poi_fused_add_div_mean_mul_std_sub_0.run(primals_2, primals_1, primals_3, buf0, 64, grid=grid(64), stream=stream0)
del primals_2
del primals_3
buf1 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(buf0, (16, 4), (4, 1), 0), reinterpret_tensor(primals_4, (4, 4), (1, 4), 0), out=buf1)
buf2 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(buf0, (16, 4), (4, 1), 0), reinterpret_tensor(primals_6, (4, 4), (1, 4), 0), out=buf2)
buf3 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(buf0, (16, 4), (4, 1), 0), reinterpret_tensor(primals_8, (4, 4), (1, 4), 0), out=buf3)
buf4 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 1), torch.float32)
# Topologically Sorted Source Nodes: [query_up_1, scores], Original ATen: [aten.div, aten.clone]
triton_poi_fused_clone_div_1.run(buf3, primals_9, buf4, 16, 4, grid=grid(16, 4), stream=stream0)
del primals_9
buf5 = reinterpret_tensor(buf3, (4, 4, 1, 4), (16, 4, 4, 1), 0); del buf3 # reuse
# Topologically Sorted Source Nodes: [scores], Original ATen: [aten.clone]
triton_poi_fused_clone_2.run(buf1, primals_5, buf5, 16, 4, grid=grid(16, 4), stream=stream0)
del primals_5
buf6 = empty_strided_cuda((16, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [scores], Original ATen: [aten.bmm]
extern_kernels.bmm(reinterpret_tensor(buf4, (16, 4, 1), (4, 1, 0), 0), reinterpret_tensor(buf5, (16, 1, 4), (4, 0, 1), 0), out=buf6)
buf7 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [attn], Original ATen: [aten._softmax]
triton_poi_fused__softmax_3.run(buf6, buf7, 256, grid=grid(256), stream=stream0)
buf8 = reinterpret_tensor(buf6, (4, 4, 4, 4), (64, 16, 4, 1), 0); del buf6 # reuse
# Topologically Sorted Source Nodes: [attn], Original ATen: [aten._softmax]
triton_poi_fused__softmax_4.run(buf7, buf8, 256, grid=grid(256), stream=stream0)
del buf7
buf9 = reinterpret_tensor(buf1, (4, 4, 4, 1), (16, 4, 1, 1), 0); del buf1 # reuse
# Topologically Sorted Source Nodes: [matmul_1], Original ATen: [aten.clone]
triton_poi_fused_clone_2.run(buf2, primals_7, buf9, 16, 4, grid=grid(16, 4), stream=stream0)
del primals_7
buf10 = reinterpret_tensor(buf2, (16, 4, 1), (4, 1, 1), 0); del buf2 # reuse
# Topologically Sorted Source Nodes: [matmul_1], Original ATen: [aten.bmm]
extern_kernels.bmm(reinterpret_tensor(buf8, (16, 4, 4), (16, 4, 1), 0), reinterpret_tensor(buf9, (16, 4, 1), (4, 1, 0), 0), out=buf10)
buf11 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 1), torch.float32)
# Topologically Sorted Source Nodes: [contiguous], Original ATen: [aten.clone]
triton_poi_fused_clone_5.run(buf10, buf11, 16, 4, grid=grid(16, 4), stream=stream0)
buf12 = reinterpret_tensor(buf10, (16, 4), (4, 1), 0); del buf10 # reuse
# Topologically Sorted Source Nodes: [output], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_11, reinterpret_tensor(buf11, (16, 4), (4, 1), 0), reinterpret_tensor(primals_10, (4, 4), (1, 4), 0), alpha=1, beta=1, out=buf12)
del primals_11
buf13 = empty_strided_cuda((4, 4, 1), (4, 1, 16), torch.float32)
buf14 = buf13; del buf13 # reuse
buf15 = empty_strided_cuda((4, 4, 1), (4, 1, 16), torch.float32)
# Topologically Sorted Source Nodes: [out, mean_2, std_2], Original ATen: [aten.add, aten.mean, aten.std]
triton_poi_fused_add_mean_std_6.run(buf14, buf12, primals_1, buf15, 16, grid=grid(16), stream=stream0)
buf16 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [out, mean_2, std_2, sub_1, mul_1, add_3, truediv_2, add_4], Original ATen: [aten.add, aten.mean, aten.std, aten.sub, aten.mul, aten.div]
triton_poi_fused_add_div_mean_mul_std_sub_7.run(primals_12, buf12, primals_1, buf15, buf14, primals_13, buf16, 64, grid=grid(64), stream=stream0)
del buf14
del buf15
del primals_13
buf17 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(buf16, (16, 4), (4, 1), 0), reinterpret_tensor(primals_14, (4, 4), (1, 4), 0), out=buf17)
buf18 = reinterpret_tensor(buf17, (4, 4, 4), (16, 4, 1), 0); del buf17 # reuse
buf22 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.bool)
# Topologically Sorted Source Nodes: [relu], Original ATen: [aten.relu, aten.threshold_backward]
triton_poi_fused_relu_threshold_backward_8.run(buf18, primals_15, buf22, 64, grid=grid(64), stream=stream0)
del primals_15
buf19 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [linear_5], Original ATen: [aten.view]
triton_poi_fused_view_9.run(buf18, buf19, 64, grid=grid(64), stream=stream0)
buf20 = reinterpret_tensor(buf18, (16, 4), (4, 1), 0); del buf18 # reuse
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(buf19, reinterpret_tensor(primals_16, (4, 4), (1, 4), 0), out=buf20)
buf21 = reinterpret_tensor(buf20, (4, 4, 4), (16, 4, 1), 0); del buf20 # reuse
# Topologically Sorted Source Nodes: [out, add_5], Original ATen: [aten.add]
triton_poi_fused_add_10.run(buf21, primals_17, buf12, primals_1, 64, grid=grid(64), stream=stream0)
del primals_17
return (buf21, primals_1, primals_12, reinterpret_tensor(buf0, (16, 4), (4, 1), 0), buf8, reinterpret_tensor(buf11, (16, 4), (4, 1), 0), buf12, reinterpret_tensor(buf16, (16, 4), (4, 1), 0), buf19, primals_16, buf22, primals_14, primals_10, reinterpret_tensor(buf9, (16, 1, 4), (4, 1, 1), 0), reinterpret_tensor(buf4, (16, 1, 4), (4, 1, 1), 0), reinterpret_tensor(buf5, (16, 4, 1), (4, 1, 4), 0), primals_8, primals_6, primals_4, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_8 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_9 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_10 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_11 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_12 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_13 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_14 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_15 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_16 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_17 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9, primals_10, primals_11, primals_12, primals_13, primals_14, primals_15, primals_16, primals_17])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import math
import torch
from torch.autograd import Variable
import torch.nn as nn
import torch.optim
class LayerNorm(nn.Module):
def __init__(self, features, eps=1e-06):
super(LayerNorm, self).__init__()
self.a_2 = nn.Parameter(torch.ones(features))
self.b_2 = nn.Parameter(torch.zeros(features))
self.eps = eps
def forward(self, x):
mean = x.mean(-1, keepdim=True)
std = x.std(-1, keepdim=True)
return self.a_2 * (x - mean) / (std + self.eps) + self.b_2
class MultiHeadedAttention(nn.Module):
"""
Multi-Head Attention module from
"Attention is All You Need"
:cite:`DBLP:journals/corr/VaswaniSPUJGKP17`.
Similar to standard `dot` attention but uses
multiple attention distributions simulataneously
to select relevant items.
.. mermaid::
graph BT
A[key]
B[value]
C[query]
O[output]
subgraph Attn
D[Attn 1]
E[Attn 2]
F[Attn N]
end
A --> D
C --> D
A --> E
C --> E
A --> F
C --> F
D --> O
E --> O
F --> O
B --> O
Also includes several additional tricks.
Args:
head_count (int): number of parallel heads
model_dim (int): the dimension of keys/values/queries,
must be divisible by head_count
dropout (float): dropout parameter
"""
def __init__(self, head_count, model_dim, dropout=0.1):
assert model_dim % head_count == 0
self.dim_per_head = model_dim // head_count
self.model_dim = model_dim
super(MultiHeadedAttention, self).__init__()
self.head_count = head_count
self.linear_keys = nn.Linear(model_dim, head_count * self.dim_per_head)
self.linear_values = nn.Linear(model_dim, head_count * self.
dim_per_head)
self.linear_query = nn.Linear(model_dim, head_count * self.dim_per_head
)
self.sm = nn.Softmax(dim=-1)
self.dropout = nn.Dropout(dropout)
self.final_linear = nn.Linear(model_dim, model_dim)
def _split_heads(self, x):
batch_size = x.size(0)
return x.view(batch_size, -1, self.head_count, self.dim_per_head
).transpose(1, 2)
def _combine_heads(self, x):
""":param x: [batch_size * head_count, seq_len, dim_per_head]"""
seq_len = x.size(2)
return x.transpose(1, 2).contiguous().view(-1, seq_len, self.
head_count * self.dim_per_head)
def compute_cache(self, key, value):
key_up = self._split_heads(self.linear_keys(key))
value_up = self._split_heads(self.linear_values(value))
return [key_up, value_up]
def forward(self, key, value, query, mask=None, cache=None):
"""
Compute the context vector and the attention vectors.
Args:
key (`FloatTensor`): set of `key_len`
key vectors `[batch, key_len, dim]`
value (`FloatTensor`): set of `key_len`
value vectors `[batch, key_len, dim]`
query (`FloatTensor`): set of `query_len`
query vectors `[batch, query_len, dim]`
mask: binary mask indicating which keys have
non-zero attention `[batch, query_len, key_len]`
Returns:
(`FloatTensor`, `FloatTensor`) :
* output context vectors `[batch, query_len, dim]`
* one of the attention vectors `[batch, query_len, key_len]`
"""
batch_size = key.size(0)
dim_per_head = self.dim_per_head
head_count = self.head_count
key_len = key.size(1)
query_len = query.size(1)
if cache is not None:
key_up, value_up = cache
else:
key_up = self._split_heads(self.linear_keys(key))
value_up = self._split_heads(self.linear_values(value))
query_up = self._split_heads(self.linear_query(query))
query_up = query_up / math.sqrt(dim_per_head)
scores = torch.matmul(query_up, key_up.transpose(2, 3))
if mask is not None:
mask = mask.unsqueeze(1).expand_as(scores)
scores = scores.masked_fill(Variable(mask), -1e+18)
attn = self.sm(scores)
drop_attn = self.dropout(attn)
context = self._combine_heads(torch.matmul(drop_attn, value_up))
output = self.final_linear(context)
top_attn = attn.view(batch_size, head_count, query_len, key_len)[:,
0, :, :].contiguous()
return output, top_attn
class PositionwiseFeedForward(nn.Module):
""" A two-layer Feed-Forward-Network with residual layer norm.
Args:
size (int): the size of input for the first-layer of the FFN.
hidden_size (int): the hidden layer size of the second-layer
of the FNN.
dropout (float): dropout probability(0-1.0).
"""
def __init__(self, size, hidden_size, dropout=0.1):
super(PositionwiseFeedForward, self).__init__()
self.w_1 = nn.Linear(size, hidden_size)
self.w_2 = nn.Linear(hidden_size, size)
self.layer_norm = LayerNorm(size)
self.dropout_1 = nn.Dropout(dropout, inplace=True)
self.relu = nn.ReLU(inplace=True)
self.dropout_2 = nn.Dropout(dropout)
def forward(self, x):
inter = self.dropout_1(self.relu(self.w_1(self.layer_norm(x))))
output = self.dropout_2(self.w_2(inter))
return output + x
class EncoderBlock(nn.Module):
def __init__(self, d_model, d_inner_hid, n_head, dropout=0.1):
super(EncoderBlock, self).__init__()
self.layer_norm = LayerNorm(features=d_model)
self.slf_attn = MultiHeadedAttention(head_count=n_head, model_dim=
d_model, dropout=dropout)
self.pos_ffn = PositionwiseFeedForward(size=d_model, hidden_size=
d_inner_hid, dropout=dropout)
self.dropout = nn.Dropout(dropout)
def forward(self, enc_input, slf_attn_mask=None):
input_norm = self.layer_norm(enc_input)
context, _ = self.slf_attn(input_norm, input_norm, input_norm,
slf_attn_mask)
out = self.dropout(context) + enc_input
return self.pos_ffn(out)
def get_inputs():
return [torch.rand([4, 4, 4])]
def get_init_inputs():
return [[], {'d_model': 4, 'd_inner_hid': 4, 'n_head': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
import math
from torch.autograd import Variable
import torch.nn as nn
import torch.optim
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_add_div_mean_mul_std_sub_0(in_ptr0, in_ptr1, in_ptr2,
out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 4
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + x2, xmask)
tmp2 = tl.load(in_ptr1 + 4 * x1, xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr1 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr1 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr1 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp30 = tl.load(in_ptr2 + x0, xmask, eviction_policy='evict_last')
tmp4 = tmp2 + tmp3
tmp6 = tmp4 + tmp5
tmp8 = tmp6 + tmp7
tmp9 = 4.0
tmp10 = tmp8 / tmp9
tmp11 = tmp1 - tmp10
tmp12 = tmp0 * tmp11
tmp13 = tmp2 - tmp10
tmp14 = tmp13 * tmp13
tmp15 = tmp3 - tmp10
tmp16 = tmp15 * tmp15
tmp17 = tmp14 + tmp16
tmp18 = tmp5 - tmp10
tmp19 = tmp18 * tmp18
tmp20 = tmp17 + tmp19
tmp21 = tmp7 - tmp10
tmp22 = tmp21 * tmp21
tmp23 = tmp20 + tmp22
tmp24 = 3.0
tmp25 = tmp23 / tmp24
tmp26 = libdevice.sqrt(tmp25)
tmp27 = 1e-06
tmp28 = tmp26 + tmp27
tmp29 = tmp12 / tmp28
tmp31 = tmp29 + tmp30
tl.store(out_ptr0 + x2, tmp31, xmask)
@triton.jit
def triton_poi_fused_clone_div_1(in_ptr0, in_ptr1, out_ptr0, ynumel, xnumel,
YBLOCK: tl.constexpr, XBLOCK: tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = yindex // 4
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + 4 * x2 + 16 * y1), xmask & ymask,
eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + y0, ymask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = 1.0
tmp4 = tmp2 * tmp3
tl.store(out_ptr0 + (x2 + 4 * y3), tmp4, xmask & ymask)
@triton.jit
def triton_poi_fused_clone_2(in_ptr0, in_ptr1, out_ptr0, ynumel, xnumel,
YBLOCK: tl.constexpr, XBLOCK: tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = yindex // 4
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + 4 * x2 + 16 * y1), xmask & ymask,
eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + y0, ymask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(out_ptr0 + (x2 + 4 * y3), tmp2, xmask & ymask)
@triton.jit
def triton_poi_fused__softmax_3(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp3 = triton_helpers.maximum(tmp1, tmp2)
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp7 = triton_helpers.maximum(tmp5, tmp6)
tmp8 = tmp0 - tmp7
tmp9 = tl_math.exp(tmp8)
tl.store(out_ptr0 + x2, tmp9, xmask)
@triton.jit
def triton_poi_fused__softmax_4(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + x2, tmp8, xmask)
@triton.jit
def triton_poi_fused_clone_5(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK: tl.
constexpr, XBLOCK: tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = yindex // 4
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + 4 * x2 + 16 * y1), xmask & ymask,
eviction_policy='evict_last')
tl.store(out_ptr0 + (x2 + 4 * y3), tmp0, xmask & ymask)
@triton.jit
def triton_poi_fused_add_mean_std_6(in_out_ptr0, in_ptr0, in_ptr1, out_ptr0,
xnumel, XBLOCK: tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + 4 * x0, xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + 4 * x0, xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr0 + (1 + 4 * x0), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr1 + (1 + 4 * x0), xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr0 + (2 + 4 * x0), xmask, eviction_policy='evict_last')
tmp8 = tl.load(in_ptr1 + (2 + 4 * x0), xmask, eviction_policy='evict_last')
tmp11 = tl.load(in_ptr0 + (3 + 4 * x0), xmask, eviction_policy='evict_last'
)
tmp12 = tl.load(in_ptr1 + (3 + 4 * x0), xmask, eviction_policy='evict_last'
)
tmp2 = tmp0 + tmp1
tmp5 = tmp3 + tmp4
tmp6 = tmp2 + tmp5
tmp9 = tmp7 + tmp8
tmp10 = tmp6 + tmp9
tmp13 = tmp11 + tmp12
tmp14 = tmp10 + tmp13
tmp15 = 4.0
tmp16 = tmp14 / tmp15
tmp17 = tmp2 - tmp16
tmp18 = tmp17 * tmp17
tmp19 = tmp5 - tmp16
tmp20 = tmp19 * tmp19
tmp21 = tmp18 + tmp20
tmp22 = tmp9 - tmp16
tmp23 = tmp22 * tmp22
tmp24 = tmp21 + tmp23
tmp25 = tmp13 - tmp16
tmp26 = tmp25 * tmp25
tmp27 = tmp24 + tmp26
tmp28 = 3.0
tmp29 = tmp27 / tmp28
tl.store(in_out_ptr0 + x0, tmp29, xmask)
tl.store(out_ptr0 + x0, tmp16, xmask)
@triton.jit
def triton_poi_fused_add_div_mean_mul_std_sub_7(in_ptr0, in_ptr1, in_ptr2,
in_ptr3, in_ptr4, in_ptr5, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 4
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + x2, xmask)
tmp2 = tl.load(in_ptr2 + x2, xmask)
tmp4 = tl.load(in_ptr3 + x1, xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr4 + x1, xmask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr5 + x0, xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 - tmp4
tmp6 = tmp0 * tmp5
tmp8 = libdevice.sqrt(tmp7)
tmp9 = 1e-06
tmp10 = tmp8 + tmp9
tmp11 = tmp6 / tmp10
tmp13 = tmp11 + tmp12
tl.store(out_ptr0 + x2, tmp13, xmask)
@triton.jit
def triton_poi_fused_relu_threshold_backward_8(in_out_ptr0, in_ptr0,
out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(in_out_ptr0 + x2, tmp4, xmask)
tl.store(out_ptr0 + x2, tmp6, xmask)
@triton.jit
def triton_poi_fused_view_9(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 4
x1 = xindex // 4
x2 = xindex
tmp0 = tl.load(in_ptr0 + (x0 + 4 * x1 + 16 * (x1 % 4 // 4)), xmask)
tl.store(out_ptr0 + x2, tmp0, xmask)
@triton.jit
def triton_poi_fused_add_10(in_out_ptr0, in_ptr0, in_ptr1, in_ptr2, xnumel,
XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr1 + x2, xmask)
tmp4 = tl.load(in_ptr2 + x2, xmask)
tmp2 = tmp0 + tmp1
tmp5 = tmp3 + tmp4
tmp6 = tmp2 + tmp5
tl.store(in_out_ptr0 + x2, tmp6, xmask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7, primals_8, primals_9, primals_10, primals_11, primals_12,
primals_13, primals_14, primals_15, primals_16, primals_17) = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_2, (4,), (1,))
assert_size_stride(primals_3, (4,), (1,))
assert_size_stride(primals_4, (4, 4), (4, 1))
assert_size_stride(primals_5, (4,), (1,))
assert_size_stride(primals_6, (4, 4), (4, 1))
assert_size_stride(primals_7, (4,), (1,))
assert_size_stride(primals_8, (4, 4), (4, 1))
assert_size_stride(primals_9, (4,), (1,))
assert_size_stride(primals_10, (4, 4), (4, 1))
assert_size_stride(primals_11, (4,), (1,))
assert_size_stride(primals_12, (4,), (1,))
assert_size_stride(primals_13, (4,), (1,))
assert_size_stride(primals_14, (4, 4), (4, 1))
assert_size_stride(primals_15, (4,), (1,))
assert_size_stride(primals_16, (4, 4), (4, 1))
assert_size_stride(primals_17, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_add_div_mean_mul_std_sub_0[grid(64)](primals_2,
primals_1, primals_3, buf0, 64, XBLOCK=64, num_warps=1,
num_stages=1)
del primals_2
del primals_3
buf1 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(buf0, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_4, (4, 4), (1, 4), 0), out=buf1)
buf2 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(buf0, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_6, (4, 4), (1, 4), 0), out=buf2)
buf3 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(buf0, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_8, (4, 4), (1, 4), 0), out=buf3)
buf4 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 1), torch.float32)
triton_poi_fused_clone_div_1[grid(16, 4)](buf3, primals_9, buf4, 16,
4, XBLOCK=2, YBLOCK=16, num_warps=1, num_stages=1)
del primals_9
buf5 = reinterpret_tensor(buf3, (4, 4, 1, 4), (16, 4, 4, 1), 0)
del buf3
triton_poi_fused_clone_2[grid(16, 4)](buf1, primals_5, buf5, 16, 4,
XBLOCK=4, YBLOCK=8, num_warps=1, num_stages=1)
del primals_5
buf6 = empty_strided_cuda((16, 4, 4), (16, 4, 1), torch.float32)
extern_kernels.bmm(reinterpret_tensor(buf4, (16, 4, 1), (4, 1, 0),
0), reinterpret_tensor(buf5, (16, 1, 4), (4, 0, 1), 0), out=buf6)
buf7 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
triton_poi_fused__softmax_3[grid(256)](buf6, buf7, 256, XBLOCK=128,
num_warps=4, num_stages=1)
buf8 = reinterpret_tensor(buf6, (4, 4, 4, 4), (64, 16, 4, 1), 0)
del buf6
triton_poi_fused__softmax_4[grid(256)](buf7, buf8, 256, XBLOCK=128,
num_warps=4, num_stages=1)
del buf7
buf9 = reinterpret_tensor(buf1, (4, 4, 4, 1), (16, 4, 1, 1), 0)
del buf1
triton_poi_fused_clone_2[grid(16, 4)](buf2, primals_7, buf9, 16, 4,
XBLOCK=4, YBLOCK=8, num_warps=1, num_stages=1)
del primals_7
buf10 = reinterpret_tensor(buf2, (16, 4, 1), (4, 1, 1), 0)
del buf2
extern_kernels.bmm(reinterpret_tensor(buf8, (16, 4, 4), (16, 4, 1),
0), reinterpret_tensor(buf9, (16, 4, 1), (4, 1, 0), 0), out=buf10)
buf11 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 1), torch.float32)
triton_poi_fused_clone_5[grid(16, 4)](buf10, buf11, 16, 4, XBLOCK=4,
YBLOCK=16, num_warps=1, num_stages=1)
buf12 = reinterpret_tensor(buf10, (16, 4), (4, 1), 0)
del buf10
extern_kernels.addmm(primals_11, reinterpret_tensor(buf11, (16, 4),
(4, 1), 0), reinterpret_tensor(primals_10, (4, 4), (1, 4), 0),
alpha=1, beta=1, out=buf12)
del primals_11
buf13 = empty_strided_cuda((4, 4, 1), (4, 1, 16), torch.float32)
buf14 = buf13
del buf13
buf15 = empty_strided_cuda((4, 4, 1), (4, 1, 16), torch.float32)
triton_poi_fused_add_mean_std_6[grid(16)](buf14, buf12, primals_1,
buf15, 16, XBLOCK=16, num_warps=1, num_stages=1)
buf16 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
triton_poi_fused_add_div_mean_mul_std_sub_7[grid(64)](primals_12,
buf12, primals_1, buf15, buf14, primals_13, buf16, 64, XBLOCK=
64, num_warps=1, num_stages=1)
del buf14
del buf15
del primals_13
buf17 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(buf16, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_14, (4, 4), (1, 4), 0), out=buf17)
buf18 = reinterpret_tensor(buf17, (4, 4, 4), (16, 4, 1), 0)
del buf17
buf22 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.bool)
triton_poi_fused_relu_threshold_backward_8[grid(64)](buf18,
primals_15, buf22, 64, XBLOCK=64, num_warps=1, num_stages=1)
del primals_15
buf19 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
triton_poi_fused_view_9[grid(64)](buf18, buf19, 64, XBLOCK=64,
num_warps=1, num_stages=1)
buf20 = reinterpret_tensor(buf18, (16, 4), (4, 1), 0)
del buf18
extern_kernels.mm(buf19, reinterpret_tensor(primals_16, (4, 4), (1,
4), 0), out=buf20)
buf21 = reinterpret_tensor(buf20, (4, 4, 4), (16, 4, 1), 0)
del buf20
triton_poi_fused_add_10[grid(64)](buf21, primals_17, buf12,
primals_1, 64, XBLOCK=64, num_warps=1, num_stages=1)
del primals_17
return buf21, primals_1, primals_12, reinterpret_tensor(buf0, (16, 4),
(4, 1), 0), buf8, reinterpret_tensor(buf11, (16, 4), (4, 1), 0
), buf12, reinterpret_tensor(buf16, (16, 4), (4, 1), 0
), buf19, primals_16, buf22, primals_14, primals_10, reinterpret_tensor(
buf9, (16, 1, 4), (4, 1, 1), 0), reinterpret_tensor(buf4, (16, 1, 4
), (4, 1, 1), 0), reinterpret_tensor(buf5, (16, 4, 1), (4, 1, 4), 0
), primals_8, primals_6, primals_4
class LayerNorm(nn.Module):
def __init__(self, features, eps=1e-06):
super(LayerNorm, self).__init__()
self.a_2 = nn.Parameter(torch.ones(features))
self.b_2 = nn.Parameter(torch.zeros(features))
self.eps = eps
def forward(self, x):
mean = x.mean(-1, keepdim=True)
std = x.std(-1, keepdim=True)
return self.a_2 * (x - mean) / (std + self.eps) + self.b_2
class MultiHeadedAttention(nn.Module):
"""
Multi-Head Attention module from
"Attention is All You Need"
:cite:`DBLP:journals/corr/VaswaniSPUJGKP17`.
Similar to standard `dot` attention but uses
multiple attention distributions simulataneously
to select relevant items.
.. mermaid::
graph BT
A[key]
B[value]
C[query]
O[output]
subgraph Attn
D[Attn 1]
E[Attn 2]
F[Attn N]
end
A --> D
C --> D
A --> E
C --> E
A --> F
C --> F
D --> O
E --> O
F --> O
B --> O
Also includes several additional tricks.
Args:
head_count (int): number of parallel heads
model_dim (int): the dimension of keys/values/queries,
must be divisible by head_count
dropout (float): dropout parameter
"""
def __init__(self, head_count, model_dim, dropout=0.1):
assert model_dim % head_count == 0
self.dim_per_head = model_dim // head_count
self.model_dim = model_dim
super(MultiHeadedAttention, self).__init__()
self.head_count = head_count
self.linear_keys = nn.Linear(model_dim, head_count * self.dim_per_head)
self.linear_values = nn.Linear(model_dim, head_count * self.
dim_per_head)
self.linear_query = nn.Linear(model_dim, head_count * self.dim_per_head
)
self.sm = nn.Softmax(dim=-1)
self.dropout = nn.Dropout(dropout)
self.final_linear = nn.Linear(model_dim, model_dim)
def _split_heads(self, x):
batch_size = x.size(0)
return x.view(batch_size, -1, self.head_count, self.dim_per_head
).transpose(1, 2)
def _combine_heads(self, x):
""":param x: [batch_size * head_count, seq_len, dim_per_head]"""
seq_len = x.size(2)
return x.transpose(1, 2).contiguous().view(-1, seq_len, self.
head_count * self.dim_per_head)
def compute_cache(self, key, value):
key_up = self._split_heads(self.linear_keys(key))
value_up = self._split_heads(self.linear_values(value))
return [key_up, value_up]
def forward(self, key, value, query, mask=None, cache=None):
"""
Compute the context vector and the attention vectors.
Args:
key (`FloatTensor`): set of `key_len`
key vectors `[batch, key_len, dim]`
value (`FloatTensor`): set of `key_len`
value vectors `[batch, key_len, dim]`
query (`FloatTensor`): set of `query_len`
query vectors `[batch, query_len, dim]`
mask: binary mask indicating which keys have
non-zero attention `[batch, query_len, key_len]`
Returns:
(`FloatTensor`, `FloatTensor`) :
* output context vectors `[batch, query_len, dim]`
* one of the attention vectors `[batch, query_len, key_len]`
"""
batch_size = key.size(0)
dim_per_head = self.dim_per_head
head_count = self.head_count
key_len = key.size(1)
query_len = query.size(1)
if cache is not None:
key_up, value_up = cache
else:
key_up = self._split_heads(self.linear_keys(key))
value_up = self._split_heads(self.linear_values(value))
query_up = self._split_heads(self.linear_query(query))
query_up = query_up / math.sqrt(dim_per_head)
scores = torch.matmul(query_up, key_up.transpose(2, 3))
if mask is not None:
mask = mask.unsqueeze(1).expand_as(scores)
scores = scores.masked_fill(Variable(mask), -1e+18)
attn = self.sm(scores)
drop_attn = self.dropout(attn)
context = self._combine_heads(torch.matmul(drop_attn, value_up))
output = self.final_linear(context)
top_attn = attn.view(batch_size, head_count, query_len, key_len)[:,
0, :, :].contiguous()
return output, top_attn
class PositionwiseFeedForward(nn.Module):
""" A two-layer Feed-Forward-Network with residual layer norm.
Args:
size (int): the size of input for the first-layer of the FFN.
hidden_size (int): the hidden layer size of the second-layer
of the FNN.
dropout (float): dropout probability(0-1.0).
"""
def __init__(self, size, hidden_size, dropout=0.1):
super(PositionwiseFeedForward, self).__init__()
self.w_1 = nn.Linear(size, hidden_size)
self.w_2 = nn.Linear(hidden_size, size)
self.layer_norm = LayerNorm(size)
self.dropout_1 = nn.Dropout(dropout, inplace=True)
self.relu = nn.ReLU(inplace=True)
self.dropout_2 = nn.Dropout(dropout)
def forward(self, x):
inter = self.dropout_1(self.relu(self.w_1(self.layer_norm(x))))
output = self.dropout_2(self.w_2(inter))
return output + x
class EncoderBlockNew(nn.Module):
def __init__(self, d_model, d_inner_hid, n_head, dropout=0.1):
super(EncoderBlockNew, self).__init__()
self.layer_norm = LayerNorm(features=d_model)
self.slf_attn = MultiHeadedAttention(head_count=n_head, model_dim=
d_model, dropout=dropout)
self.pos_ffn = PositionwiseFeedForward(size=d_model, hidden_size=
d_inner_hid, dropout=dropout)
self.dropout = nn.Dropout(dropout)
def forward(self, input_0):
primals_2 = self.layer_norm.a_2
primals_3 = self.layer_norm.b_2
primals_4 = self.slf_attn.linear_keys.weight
primals_5 = self.slf_attn.linear_keys.bias
primals_6 = self.slf_attn.linear_values.weight
primals_7 = self.slf_attn.linear_values.bias
primals_8 = self.slf_attn.linear_query.weight
primals_9 = self.slf_attn.linear_query.bias
primals_10 = self.slf_attn.final_linear.weight
primals_11 = self.slf_attn.final_linear.bias
primals_14 = self.pos_ffn.w_1.weight
primals_12 = self.pos_ffn.w_1.bias
primals_16 = self.pos_ffn.w_2.weight
primals_13 = self.pos_ffn.w_2.bias
primals_15 = self.pos_ffn.layer_norm.a_2
primals_17 = self.pos_ffn.layer_norm.b_2
primals_1 = input_0
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7, primals_8, primals_9,
primals_10, primals_11, primals_12, primals_13, primals_14,
primals_15, primals_16, primals_17])
return output[0]
| howardchenhd/Transformer-pytorch | EncoderBlock | false | 6,849 | [
"MIT"
] | 1 | ae71ed5767272feb7e717be6d5bfce46f80ec57a | https://github.com/howardchenhd/Transformer-pytorch/tree/ae71ed5767272feb7e717be6d5bfce46f80ec57a | import math
import torch
from torch.autograd import Variable
import torch.nn as nn
import torch.optim
class LayerNorm(nn.Module):
def __init__(self, features, eps=1e-06):
super().__init__()
self.a_2 = nn.Parameter(torch.ones(features))
self.b_2 = nn.Parameter(torch.zeros(features))
self.eps = eps
def forward(self, x):
mean = x.mean(-1, keepdim=True)
std = x.std(-1, keepdim=True)
return self.a_2 * (x - mean) / (std + self.eps) + self.b_2
class MultiHeadedAttention(nn.Module):
"""
Multi-Head Attention module from
"Attention is All You Need"
:cite:`DBLP:journals/corr/VaswaniSPUJGKP17`.
Similar to standard `dot` attention but uses
multiple attention distributions simulataneously
to select relevant items.
.. mermaid::
graph BT
A[key]
B[value]
C[query]
O[output]
subgraph Attn
D[Attn 1]
E[Attn 2]
F[Attn N]
end
A --> D
C --> D
A --> E
C --> E
A --> F
C --> F
D --> O
E --> O
F --> O
B --> O
Also includes several additional tricks.
Args:
head_count (int): number of parallel heads
model_dim (int): the dimension of keys/values/queries,
must be divisible by head_count
dropout (float): dropout parameter
"""
def __init__(self, head_count, model_dim, dropout=0.1):
assert model_dim % head_count == 0
self.dim_per_head = model_dim // head_count
self.model_dim = model_dim
super().__init__()
self.head_count = head_count
self.linear_keys = nn.Linear(model_dim, head_count * self.dim_per_head)
self.linear_values = nn.Linear(model_dim, head_count * self.
dim_per_head)
self.linear_query = nn.Linear(model_dim, head_count * self.dim_per_head
)
self.sm = nn.Softmax(dim=-1)
self.dropout = nn.Dropout(dropout)
self.final_linear = nn.Linear(model_dim, model_dim)
def _split_heads(self, x):
batch_size = x.size(0)
return x.view(batch_size, -1, self.head_count, self.dim_per_head
).transpose(1, 2)
def _combine_heads(self, x):
""":param x: [batch_size * head_count, seq_len, dim_per_head]"""
seq_len = x.size(2)
return x.transpose(1, 2).contiguous().view(-1, seq_len, self.
head_count * self.dim_per_head)
def compute_cache(self, key, value):
key_up = self._split_heads(self.linear_keys(key))
value_up = self._split_heads(self.linear_values(value))
return [key_up, value_up]
def forward(self, key, value, query, mask=None, cache=None):
"""
Compute the context vector and the attention vectors.
Args:
key (`FloatTensor`): set of `key_len`
key vectors `[batch, key_len, dim]`
value (`FloatTensor`): set of `key_len`
value vectors `[batch, key_len, dim]`
query (`FloatTensor`): set of `query_len`
query vectors `[batch, query_len, dim]`
mask: binary mask indicating which keys have
non-zero attention `[batch, query_len, key_len]`
Returns:
(`FloatTensor`, `FloatTensor`) :
* output context vectors `[batch, query_len, dim]`
* one of the attention vectors `[batch, query_len, key_len]`
"""
batch_size = key.size(0)
dim_per_head = self.dim_per_head
head_count = self.head_count
key_len = key.size(1)
query_len = query.size(1)
if cache is not None:
key_up, value_up = cache
else:
key_up = self._split_heads(self.linear_keys(key))
value_up = self._split_heads(self.linear_values(value))
query_up = self._split_heads(self.linear_query(query))
# ... truncated (>4000 chars) for memory efficiency |
DecoderBlock | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/xx/cxxsy7adcakcv5c6imnaqsvf3ebhgbph77vaoembzakbqdrjycbc.py
# Topologically Sorted Source Nodes: [mean, std, sub, mul, add, truediv, input_norm], Original ATen: [aten.mean, aten.std, aten.sub, aten.mul, aten.add, aten.div]
# Source node to ATen node mapping:
# add => add
# input_norm => add_1
# mean => mean
# mul => mul
# std => sqrt, var
# sub => sub
# truediv => div
# Graph fragment:
# %mean : [num_users=1] = call_function[target=torch.ops.aten.mean.dim](args = (%primals_1, [-1], True), kwargs = {})
# %var : [num_users=1] = call_function[target=torch.ops.aten.var.correction](args = (%primals_1, [-1]), kwargs = {correction: 1.0, keepdim: True})
# %sqrt : [num_users=1] = call_function[target=torch.ops.aten.sqrt.default](args = (%var,), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%primals_1, %mean), kwargs = {})
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%primals_3, %sub), kwargs = {})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%sqrt, 1e-06), kwargs = {})
# %div : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%mul, %add), kwargs = {})
# %add_1 : [num_users=4] = call_function[target=torch.ops.aten.add.Tensor](args = (%div, %primals_4), kwargs = {})
triton_poi_fused_add_div_mean_mul_std_sub_0 = async_compile.triton('triton_poi_fused_add_div_mean_mul_std_sub_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_div_mean_mul_std_sub_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 7, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_div_mean_mul_std_sub_0(in_ptr0, in_ptr1, in_ptr2, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 4
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (x2), xmask)
tmp2 = tl.load(in_ptr1 + (4*x1), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr1 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr1 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr1 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp30 = tl.load(in_ptr2 + (x0), xmask, eviction_policy='evict_last')
tmp4 = tmp2 + tmp3
tmp6 = tmp4 + tmp5
tmp8 = tmp6 + tmp7
tmp9 = 4.0
tmp10 = tmp8 / tmp9
tmp11 = tmp1 - tmp10
tmp12 = tmp0 * tmp11
tmp13 = tmp2 - tmp10
tmp14 = tmp13 * tmp13
tmp15 = tmp3 - tmp10
tmp16 = tmp15 * tmp15
tmp17 = tmp14 + tmp16
tmp18 = tmp5 - tmp10
tmp19 = tmp18 * tmp18
tmp20 = tmp17 + tmp19
tmp21 = tmp7 - tmp10
tmp22 = tmp21 * tmp21
tmp23 = tmp20 + tmp22
tmp24 = 3.0
tmp25 = tmp23 / tmp24
tmp26 = libdevice.sqrt(tmp25)
tmp27 = 1e-06
tmp28 = tmp26 + tmp27
tmp29 = tmp12 / tmp28
tmp31 = tmp29 + tmp30
tl.store(out_ptr0 + (x2), tmp31, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/wc/cwc2g4aer7dhpqcbpvbocvjyfzzfliu4x7goxlv7aksmsfu7j6nh.py
# Topologically Sorted Source Nodes: [query_up_1, scores], Original ATen: [aten.div, aten.clone]
# Source node to ATen node mapping:
# query_up_1 => div_1
# scores => clone
# Graph fragment:
# %div_1 : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%permute_5, 1.0), kwargs = {})
# %clone : [num_users=1] = call_function[target=torch.ops.aten.clone.default](args = (%expand,), kwargs = {memory_format: torch.contiguous_format})
triton_poi_fused_clone_div_1 = async_compile.triton('triton_poi_fused_clone_div_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16, 4], tile_hint=TileHint.DEFAULT,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clone_div_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clone_div_1(in_ptr0, in_ptr1, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = (yindex // 4)
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + (4*x2) + (16*y1)), xmask & ymask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (y0), ymask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = 1.0
tmp4 = tmp2 * tmp3
tl.store(out_ptr0 + (x2 + (4*y3)), tmp4, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/7k/c7kkxqo5r65gqykuvge3exgf3trgxmm4raf7gypitw4ynuylbeao.py
# Topologically Sorted Source Nodes: [scores], Original ATen: [aten.clone]
# Source node to ATen node mapping:
# scores => clone_1
# Graph fragment:
# %clone_1 : [num_users=1] = call_function[target=torch.ops.aten.clone.default](args = (%expand_1,), kwargs = {memory_format: torch.contiguous_format})
triton_poi_fused_clone_2 = async_compile.triton('triton_poi_fused_clone_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16, 4], tile_hint=TileHint.DEFAULT,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clone_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clone_2(in_ptr0, in_ptr1, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = (yindex // 4)
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + (4*x2) + (16*y1)), xmask & ymask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (y0), ymask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(out_ptr0 + (x2 + (4*y3)), tmp2, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/s4/cs472yivvl3yzse325afzknsz7ua5dqrqzmwls3lwujk3hte6xkl.py
# Topologically Sorted Source Nodes: [attn], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# attn => amax, exp, sub_1
# Graph fragment:
# %amax : [num_users=1] = call_function[target=torch.ops.aten.amax.default](args = (%view_11, [-1], True), kwargs = {})
# %sub_1 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%view_11, %amax), kwargs = {})
# %exp : [num_users=2] = call_function[target=torch.ops.aten.exp.default](args = (%sub_1,), kwargs = {})
triton_poi_fused__softmax_3 = async_compile.triton('triton_poi_fused__softmax_3', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_3', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_3(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp3 = triton_helpers.maximum(tmp1, tmp2)
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp7 = triton_helpers.maximum(tmp5, tmp6)
tmp8 = tmp0 - tmp7
tmp9 = tl_math.exp(tmp8)
tl.store(out_ptr0 + (x2), tmp9, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/zh/czh6tw7ngffcygnivwvcjex5edxy3ms4t27ymyn2hemxlpspxzq7.py
# Topologically Sorted Source Nodes: [attn], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# attn => div_2, sum_1
# Graph fragment:
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%exp, [-1], True), kwargs = {})
# %div_2 : [num_users=2] = call_function[target=torch.ops.aten.div.Tensor](args = (%exp, %sum_1), kwargs = {})
triton_poi_fused__softmax_4 = async_compile.triton('triton_poi_fused__softmax_4', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_4', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_4(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + (x2), tmp8, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/we/cwe54p4p4jvwbdktkpj3wy2coheu6f3r3dgvi7ozm7xjfk4mgbwx.py
# Topologically Sorted Source Nodes: [contiguous], Original ATen: [aten.clone]
# Source node to ATen node mapping:
# contiguous => clone_4
# Graph fragment:
# %clone_4 : [num_users=1] = call_function[target=torch.ops.aten.clone.default](args = (%permute_7,), kwargs = {memory_format: torch.contiguous_format})
triton_poi_fused_clone_5 = async_compile.triton('triton_poi_fused_clone_5', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16, 4], tile_hint=TileHint.SQUARE,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clone_5', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clone_5(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = (yindex // 4)
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + (4*x2) + (16*y1)), xmask & ymask, eviction_policy='evict_last')
tl.store(out_ptr0 + (x2 + (4*y3)), tmp0, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/k3/ck3fooheqiq5spsm2bt5dle5th5oggkmlao36xduvmqkrltazlaw.py
# Topologically Sorted Source Nodes: [query, mean_2, std_2], Original ATen: [aten.add, aten.mean, aten.std]
# Source node to ATen node mapping:
# mean_2 => mean_1
# query => add_2
# std_2 => var_1
# Graph fragment:
# %add_2 : [num_users=4] = call_function[target=torch.ops.aten.add.Tensor](args = (%view_17, %primals_1), kwargs = {})
# %mean_1 : [num_users=1] = call_function[target=torch.ops.aten.mean.dim](args = (%add_2, [-1], True), kwargs = {})
# %var_1 : [num_users=1] = call_function[target=torch.ops.aten.var.correction](args = (%add_2, [-1]), kwargs = {correction: 1.0, keepdim: True})
triton_poi_fused_add_mean_std_6 = async_compile.triton('triton_poi_fused_add_mean_std_6', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_mean_std_6', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 8, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_mean_std_6(in_out_ptr0, in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (4*x0), xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (4*x0), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr0 + (1 + (4*x0)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr1 + (1 + (4*x0)), xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr0 + (2 + (4*x0)), xmask, eviction_policy='evict_last')
tmp8 = tl.load(in_ptr1 + (2 + (4*x0)), xmask, eviction_policy='evict_last')
tmp11 = tl.load(in_ptr0 + (3 + (4*x0)), xmask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr1 + (3 + (4*x0)), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp5 = tmp3 + tmp4
tmp6 = tmp2 + tmp5
tmp9 = tmp7 + tmp8
tmp10 = tmp6 + tmp9
tmp13 = tmp11 + tmp12
tmp14 = tmp10 + tmp13
tmp15 = 4.0
tmp16 = tmp14 / tmp15
tmp17 = tmp2 - tmp16
tmp18 = tmp17 * tmp17
tmp19 = tmp5 - tmp16
tmp20 = tmp19 * tmp19
tmp21 = tmp18 + tmp20
tmp22 = tmp9 - tmp16
tmp23 = tmp22 * tmp22
tmp24 = tmp21 + tmp23
tmp25 = tmp13 - tmp16
tmp26 = tmp25 * tmp25
tmp27 = tmp24 + tmp26
tmp28 = 3.0
tmp29 = tmp27 / tmp28
tl.store(in_out_ptr0 + (x0), tmp29, xmask)
tl.store(out_ptr0 + (x0), tmp16, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/44/c44nsundpysjinc27mkdhx3tcd7nqj5ir3mp6sd7nqx3ryqium46.py
# Topologically Sorted Source Nodes: [query, mean_2, std_2, sub_1, mul_1, add_3, truediv_2, query_norm], Original ATen: [aten.add, aten.mean, aten.std, aten.sub, aten.mul, aten.div]
# Source node to ATen node mapping:
# add_3 => add_3
# mean_2 => mean_1
# mul_1 => mul_1
# query => add_2
# query_norm => add_4
# std_2 => sqrt_1
# sub_1 => sub_2
# truediv_2 => div_3
# Graph fragment:
# %add_2 : [num_users=4] = call_function[target=torch.ops.aten.add.Tensor](args = (%view_17, %primals_1), kwargs = {})
# %mean_1 : [num_users=1] = call_function[target=torch.ops.aten.mean.dim](args = (%add_2, [-1], True), kwargs = {})
# %sqrt_1 : [num_users=1] = call_function[target=torch.ops.aten.sqrt.default](args = (%var_1,), kwargs = {})
# %sub_2 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%add_2, %mean_1), kwargs = {})
# %mul_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%primals_13, %sub_2), kwargs = {})
# %add_3 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%sqrt_1, 1e-06), kwargs = {})
# %div_3 : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%mul_1, %add_3), kwargs = {})
# %add_4 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%div_3, %primals_14), kwargs = {})
triton_poi_fused_add_div_mean_mul_std_sub_7 = async_compile.triton('triton_poi_fused_add_div_mean_mul_std_sub_7', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: '*fp32', 5: '*fp32', 6: '*fp32', 7: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5, 6, 7), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_div_mean_mul_std_sub_7', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 6, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_div_mean_mul_std_sub_7(in_ptr0, in_ptr1, in_ptr2, in_ptr3, in_ptr4, in_ptr5, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 4
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (x2), xmask)
tmp2 = tl.load(in_ptr2 + (x2), xmask)
tmp4 = tl.load(in_ptr3 + (x1), xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr4 + (x1), xmask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr5 + (x0), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 - tmp4
tmp6 = tmp0 * tmp5
tmp8 = libdevice.sqrt(tmp7)
tmp9 = 1e-06
tmp10 = tmp8 + tmp9
tmp11 = tmp6 / tmp10
tmp13 = tmp11 + tmp12
tl.store(out_ptr0 + (x2), tmp13, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/5v/c5vntxfo7kep3uhno2zmlazyr72d6eyki3yny2wzqazwmq22k3vv.py
# Topologically Sorted Source Nodes: [top_attn_1], Original ATen: [aten.clone]
# Source node to ATen node mapping:
# top_attn_1 => clone_12
# Graph fragment:
# %clone_12 : [num_users=1] = call_function[target=torch.ops.aten.clone.default](args = (%select_1,), kwargs = {memory_format: torch.contiguous_format})
triton_poi_fused_clone_8 = async_compile.triton('triton_poi_fused_clone_8', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clone_8', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clone_8(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 16
x1 = (xindex // 16)
x2 = xindex
tmp0 = tl.load(in_ptr0 + (x0 + (64*x1)), xmask)
tl.store(out_ptr0 + (x2), tmp0, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/bx/cbxzdkbxnjj4ysq7xrysng3xgzgpomqcvqe5mdc477yiqcwvtclw.py
# Topologically Sorted Source Nodes: [query, add_5], Original ATen: [aten.add]
# Source node to ATen node mapping:
# add_5 => add_5
# query => add_2
# Graph fragment:
# %add_2 : [num_users=4] = call_function[target=torch.ops.aten.add.Tensor](args = (%view_17, %primals_1), kwargs = {})
# %add_5 : [num_users=5] = call_function[target=torch.ops.aten.add.Tensor](args = (%view_36, %add_2), kwargs = {})
triton_poi_fused_add_9 = async_compile.triton('triton_poi_fused_add_9', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_9', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 4, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_9(in_out_ptr0, in_ptr0, in_ptr1, in_ptr2, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr1 + (x2), xmask)
tmp4 = tl.load(in_ptr2 + (x2), xmask)
tmp2 = tmp0 + tmp1
tmp5 = tmp3 + tmp4
tmp6 = tmp2 + tmp5
tl.store(in_out_ptr0 + (x2), tmp6, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/ol/coljc3cs7q6hio3b326qftbftlnedhni7mopq6hnmq3reufgrywy.py
# Topologically Sorted Source Nodes: [relu], Original ATen: [aten.relu, aten.threshold_backward]
# Source node to ATen node mapping:
# relu => relu
# Graph fragment:
# %relu : [num_users=1] = call_function[target=torch.ops.aten.relu.default](args = (%view_39,), kwargs = {})
# %le : [num_users=1] = call_function[target=torch.ops.aten.le.Scalar](args = (%view_44, 0), kwargs = {})
triton_poi_fused_relu_threshold_backward_10 = async_compile.triton('triton_poi_fused_relu_threshold_backward_10', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*i1', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_relu_threshold_backward_10', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_relu_threshold_backward_10(in_out_ptr0, in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(in_out_ptr0 + (x2), tmp4, xmask)
tl.store(out_ptr0 + (x2), tmp6, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/bg/cbg7owgmfi2mnbyuzqnl62xpa7l324g7fysqnvnh2yuplhxd5hpv.py
# Topologically Sorted Source Nodes: [linear_9], Original ATen: [aten.view]
# Source node to ATen node mapping:
# linear_9 => view_45
# Graph fragment:
# %view_45 : [num_users=2] = call_function[target=torch.ops.aten.reshape.default](args = (%view_44, [16, 4]), kwargs = {})
triton_poi_fused_view_11 = async_compile.triton('triton_poi_fused_view_11', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_view_11', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_view_11(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 4
x1 = (xindex // 4)
x2 = xindex
tmp0 = tl.load(in_ptr0 + (x0 + (4*x1) + (16*((x1 % 4) // 4))), xmask)
tl.store(out_ptr0 + (x2), tmp0, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/ly/clych72evfl43h75debtcdhychoab7iw7f2ch2b5ebe3jjglsndg.py
# Topologically Sorted Source Nodes: [output_3], Original ATen: [aten.add]
# Source node to ATen node mapping:
# output_3 => add_8
# Graph fragment:
# %add_8 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%view_46, %add_5), kwargs = {})
triton_poi_fused_add_12 = async_compile.triton('triton_poi_fused_add_12', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_12', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 3, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_12(in_out_ptr0, in_ptr0, in_ptr1, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr1 + (x2), xmask)
tmp2 = tmp0 + tmp1
tmp4 = tmp2 + tmp3
tl.store(in_out_ptr0 + (x2), tmp4, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9, primals_10, primals_11, primals_12, primals_13, primals_14, primals_15, primals_16, primals_17, primals_18, primals_19, primals_20, primals_21, primals_22, primals_23, primals_24, primals_25, primals_26, primals_27, primals_28 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_2, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_3, (4, ), (1, ))
assert_size_stride(primals_4, (4, ), (1, ))
assert_size_stride(primals_5, (4, 4), (4, 1))
assert_size_stride(primals_6, (4, ), (1, ))
assert_size_stride(primals_7, (4, 4), (4, 1))
assert_size_stride(primals_8, (4, ), (1, ))
assert_size_stride(primals_9, (4, 4), (4, 1))
assert_size_stride(primals_10, (4, ), (1, ))
assert_size_stride(primals_11, (4, 4), (4, 1))
assert_size_stride(primals_12, (4, ), (1, ))
assert_size_stride(primals_13, (4, ), (1, ))
assert_size_stride(primals_14, (4, ), (1, ))
assert_size_stride(primals_15, (4, 4), (4, 1))
assert_size_stride(primals_16, (4, ), (1, ))
assert_size_stride(primals_17, (4, 4), (4, 1))
assert_size_stride(primals_18, (4, ), (1, ))
assert_size_stride(primals_19, (4, 4), (4, 1))
assert_size_stride(primals_20, (4, ), (1, ))
assert_size_stride(primals_21, (4, 4), (4, 1))
assert_size_stride(primals_22, (4, ), (1, ))
assert_size_stride(primals_23, (4, ), (1, ))
assert_size_stride(primals_24, (4, ), (1, ))
assert_size_stride(primals_25, (4, 4), (4, 1))
assert_size_stride(primals_26, (4, ), (1, ))
assert_size_stride(primals_27, (4, 4), (4, 1))
assert_size_stride(primals_28, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [mean, std, sub, mul, add, truediv, input_norm], Original ATen: [aten.mean, aten.std, aten.sub, aten.mul, aten.add, aten.div]
stream0 = get_raw_stream(0)
triton_poi_fused_add_div_mean_mul_std_sub_0.run(primals_3, primals_1, primals_4, buf0, 64, grid=grid(64), stream=stream0)
del primals_3
del primals_4
buf1 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(buf0, (16, 4), (4, 1), 0), reinterpret_tensor(primals_5, (4, 4), (1, 4), 0), out=buf1)
buf2 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(buf0, (16, 4), (4, 1), 0), reinterpret_tensor(primals_7, (4, 4), (1, 4), 0), out=buf2)
buf3 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(buf0, (16, 4), (4, 1), 0), reinterpret_tensor(primals_9, (4, 4), (1, 4), 0), out=buf3)
buf4 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 1), torch.float32)
# Topologically Sorted Source Nodes: [query_up_1, scores], Original ATen: [aten.div, aten.clone]
triton_poi_fused_clone_div_1.run(buf3, primals_10, buf4, 16, 4, grid=grid(16, 4), stream=stream0)
del primals_10
buf5 = reinterpret_tensor(buf3, (4, 4, 1, 4), (16, 4, 4, 1), 0); del buf3 # reuse
# Topologically Sorted Source Nodes: [scores], Original ATen: [aten.clone]
triton_poi_fused_clone_2.run(buf1, primals_6, buf5, 16, 4, grid=grid(16, 4), stream=stream0)
del primals_6
buf6 = empty_strided_cuda((16, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [scores], Original ATen: [aten.bmm]
extern_kernels.bmm(reinterpret_tensor(buf4, (16, 4, 1), (4, 1, 0), 0), reinterpret_tensor(buf5, (16, 1, 4), (4, 0, 1), 0), out=buf6)
buf7 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [attn], Original ATen: [aten._softmax]
triton_poi_fused__softmax_3.run(buf6, buf7, 256, grid=grid(256), stream=stream0)
buf8 = reinterpret_tensor(buf6, (4, 4, 4, 4), (64, 16, 4, 1), 0); del buf6 # reuse
# Topologically Sorted Source Nodes: [attn], Original ATen: [aten._softmax]
triton_poi_fused__softmax_4.run(buf7, buf8, 256, grid=grid(256), stream=stream0)
buf9 = reinterpret_tensor(buf1, (4, 4, 4, 1), (16, 4, 1, 1), 0); del buf1 # reuse
# Topologically Sorted Source Nodes: [matmul_1], Original ATen: [aten.clone]
triton_poi_fused_clone_2.run(buf2, primals_8, buf9, 16, 4, grid=grid(16, 4), stream=stream0)
del primals_8
buf10 = reinterpret_tensor(buf2, (16, 4, 1), (4, 1, 1), 0); del buf2 # reuse
# Topologically Sorted Source Nodes: [matmul_1], Original ATen: [aten.bmm]
extern_kernels.bmm(reinterpret_tensor(buf8, (16, 4, 4), (16, 4, 1), 0), reinterpret_tensor(buf9, (16, 4, 1), (4, 1, 0), 0), out=buf10)
buf11 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 1), torch.float32)
# Topologically Sorted Source Nodes: [contiguous], Original ATen: [aten.clone]
triton_poi_fused_clone_5.run(buf10, buf11, 16, 4, grid=grid(16, 4), stream=stream0)
buf12 = reinterpret_tensor(buf10, (16, 4), (4, 1), 0); del buf10 # reuse
# Topologically Sorted Source Nodes: [output], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_12, reinterpret_tensor(buf11, (16, 4), (4, 1), 0), reinterpret_tensor(primals_11, (4, 4), (1, 4), 0), alpha=1, beta=1, out=buf12)
del primals_12
buf13 = empty_strided_cuda((4, 4, 1), (4, 1, 16), torch.float32)
buf14 = buf13; del buf13 # reuse
buf15 = empty_strided_cuda((4, 4, 1), (4, 1, 16), torch.float32)
# Topologically Sorted Source Nodes: [query, mean_2, std_2], Original ATen: [aten.add, aten.mean, aten.std]
triton_poi_fused_add_mean_std_6.run(buf14, buf12, primals_1, buf15, 16, grid=grid(16), stream=stream0)
buf16 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(primals_2, (16, 4), (4, 1), 0), reinterpret_tensor(primals_15, (4, 4), (1, 4), 0), out=buf16)
del primals_15
buf17 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(primals_2, (16, 4), (4, 1), 0), reinterpret_tensor(primals_17, (4, 4), (1, 4), 0), out=buf17)
del primals_17
buf18 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [query, mean_2, std_2, sub_1, mul_1, add_3, truediv_2, query_norm], Original ATen: [aten.add, aten.mean, aten.std, aten.sub, aten.mul, aten.div]
triton_poi_fused_add_div_mean_mul_std_sub_7.run(primals_13, buf12, primals_1, buf15, buf14, primals_14, buf18, 64, grid=grid(64), stream=stream0)
del buf14
del buf15
del primals_14
buf19 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(buf18, (16, 4), (4, 1), 0), reinterpret_tensor(primals_19, (4, 4), (1, 4), 0), out=buf19)
buf20 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 1), torch.float32)
# Topologically Sorted Source Nodes: [query_up_3, scores_1], Original ATen: [aten.div, aten.clone]
triton_poi_fused_clone_div_1.run(buf19, primals_20, buf20, 16, 4, grid=grid(16, 4), stream=stream0)
del primals_20
buf21 = reinterpret_tensor(buf19, (4, 4, 1, 4), (16, 4, 4, 1), 0); del buf19 # reuse
# Topologically Sorted Source Nodes: [scores_1], Original ATen: [aten.clone]
triton_poi_fused_clone_2.run(buf16, primals_16, buf21, 16, 4, grid=grid(16, 4), stream=stream0)
del primals_16
buf22 = reinterpret_tensor(buf7, (16, 4, 4), (16, 4, 1), 0); del buf7 # reuse
# Topologically Sorted Source Nodes: [scores_1], Original ATen: [aten.bmm]
extern_kernels.bmm(reinterpret_tensor(buf20, (16, 4, 1), (4, 1, 0), 0), reinterpret_tensor(buf21, (16, 1, 4), (4, 0, 1), 0), out=buf22)
buf23 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [attn_1], Original ATen: [aten._softmax]
triton_poi_fused__softmax_3.run(buf22, buf23, 256, grid=grid(256), stream=stream0)
buf24 = reinterpret_tensor(buf22, (4, 4, 4, 4), (64, 16, 4, 1), 0); del buf22 # reuse
# Topologically Sorted Source Nodes: [attn_1], Original ATen: [aten._softmax]
triton_poi_fused__softmax_4.run(buf23, buf24, 256, grid=grid(256), stream=stream0)
del buf23
buf25 = reinterpret_tensor(buf16, (4, 4, 4, 1), (16, 4, 1, 1), 0); del buf16 # reuse
# Topologically Sorted Source Nodes: [matmul_3], Original ATen: [aten.clone]
triton_poi_fused_clone_2.run(buf17, primals_18, buf25, 16, 4, grid=grid(16, 4), stream=stream0)
del primals_18
buf26 = reinterpret_tensor(buf17, (16, 4, 1), (4, 1, 1), 0); del buf17 # reuse
# Topologically Sorted Source Nodes: [matmul_3], Original ATen: [aten.bmm]
extern_kernels.bmm(reinterpret_tensor(buf24, (16, 4, 4), (16, 4, 1), 0), reinterpret_tensor(buf25, (16, 4, 1), (4, 1, 0), 0), out=buf26)
buf27 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 1), torch.float32)
# Topologically Sorted Source Nodes: [contiguous_2], Original ATen: [aten.clone]
triton_poi_fused_clone_5.run(buf26, buf27, 16, 4, grid=grid(16, 4), stream=stream0)
buf28 = reinterpret_tensor(buf26, (16, 4), (4, 1), 0); del buf26 # reuse
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(buf27, (16, 4), (4, 1), 0), reinterpret_tensor(primals_21, (4, 4), (1, 4), 0), out=buf28)
buf29 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [top_attn_1], Original ATen: [aten.clone]
triton_poi_fused_clone_8.run(buf24, buf29, 64, grid=grid(64), stream=stream0)
buf30 = reinterpret_tensor(buf28, (4, 4, 4), (16, 4, 1), 0); del buf28 # reuse
# Topologically Sorted Source Nodes: [query, add_5], Original ATen: [aten.add]
triton_poi_fused_add_9.run(buf30, primals_22, buf12, primals_1, 64, grid=grid(64), stream=stream0)
del primals_22
buf31 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [mean_4, std_4, sub_2, mul_2, add_6, truediv_4, add_7], Original ATen: [aten.mean, aten.std, aten.sub, aten.mul, aten.add, aten.div]
triton_poi_fused_add_div_mean_mul_std_sub_0.run(primals_23, buf30, primals_24, buf31, 64, grid=grid(64), stream=stream0)
del primals_24
buf32 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(buf31, (16, 4), (4, 1), 0), reinterpret_tensor(primals_25, (4, 4), (1, 4), 0), out=buf32)
buf33 = reinterpret_tensor(buf32, (4, 4, 4), (16, 4, 1), 0); del buf32 # reuse
buf37 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.bool)
# Topologically Sorted Source Nodes: [relu], Original ATen: [aten.relu, aten.threshold_backward]
triton_poi_fused_relu_threshold_backward_10.run(buf33, primals_26, buf37, 64, grid=grid(64), stream=stream0)
del primals_26
buf34 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [linear_9], Original ATen: [aten.view]
triton_poi_fused_view_11.run(buf33, buf34, 64, grid=grid(64), stream=stream0)
buf35 = reinterpret_tensor(buf33, (16, 4), (4, 1), 0); del buf33 # reuse
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(buf34, reinterpret_tensor(primals_27, (4, 4), (1, 4), 0), out=buf35)
buf36 = reinterpret_tensor(buf35, (4, 4, 4), (16, 4, 1), 0); del buf35 # reuse
# Topologically Sorted Source Nodes: [output_3], Original ATen: [aten.add]
triton_poi_fused_add_12.run(buf36, primals_28, buf30, 64, grid=grid(64), stream=stream0)
del primals_28
return (buf36, buf29, buf0, primals_1, primals_13, primals_23, buf0, buf8, reinterpret_tensor(buf11, (16, 4), (4, 1), 0), buf12, reinterpret_tensor(primals_2, (16, 4), (4, 1), 0), reinterpret_tensor(buf18, (16, 4), (4, 1), 0), buf24, reinterpret_tensor(buf27, (16, 4), (4, 1), 0), buf30, reinterpret_tensor(buf31, (16, 4), (4, 1), 0), buf34, primals_27, buf37, primals_25, primals_21, reinterpret_tensor(buf25, (16, 1, 4), (4, 1, 1), 0), reinterpret_tensor(buf20, (16, 1, 4), (4, 1, 1), 0), reinterpret_tensor(buf21, (16, 4, 1), (4, 1, 4), 0), primals_19, primals_11, reinterpret_tensor(buf9, (16, 1, 4), (4, 1, 1), 0), reinterpret_tensor(buf4, (16, 1, 4), (4, 1, 1), 0), reinterpret_tensor(buf5, (16, 4, 1), (4, 1, 4), 0), primals_9, primals_7, primals_5, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_8 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_9 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_10 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_11 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_12 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_13 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_14 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_15 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_16 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_17 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_18 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_19 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_20 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_21 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_22 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_23 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_24 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_25 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_26 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_27 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_28 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9, primals_10, primals_11, primals_12, primals_13, primals_14, primals_15, primals_16, primals_17, primals_18, primals_19, primals_20, primals_21, primals_22, primals_23, primals_24, primals_25, primals_26, primals_27, primals_28])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import math
import torch
from torch.autograd import Variable
import torch.nn as nn
import torch.optim
class LayerNorm(nn.Module):
def __init__(self, features, eps=1e-06):
super(LayerNorm, self).__init__()
self.a_2 = nn.Parameter(torch.ones(features))
self.b_2 = nn.Parameter(torch.zeros(features))
self.eps = eps
def forward(self, x):
mean = x.mean(-1, keepdim=True)
std = x.std(-1, keepdim=True)
return self.a_2 * (x - mean) / (std + self.eps) + self.b_2
class MultiHeadedAttention(nn.Module):
"""
Multi-Head Attention module from
"Attention is All You Need"
:cite:`DBLP:journals/corr/VaswaniSPUJGKP17`.
Similar to standard `dot` attention but uses
multiple attention distributions simulataneously
to select relevant items.
.. mermaid::
graph BT
A[key]
B[value]
C[query]
O[output]
subgraph Attn
D[Attn 1]
E[Attn 2]
F[Attn N]
end
A --> D
C --> D
A --> E
C --> E
A --> F
C --> F
D --> O
E --> O
F --> O
B --> O
Also includes several additional tricks.
Args:
head_count (int): number of parallel heads
model_dim (int): the dimension of keys/values/queries,
must be divisible by head_count
dropout (float): dropout parameter
"""
def __init__(self, head_count, model_dim, dropout=0.1):
assert model_dim % head_count == 0
self.dim_per_head = model_dim // head_count
self.model_dim = model_dim
super(MultiHeadedAttention, self).__init__()
self.head_count = head_count
self.linear_keys = nn.Linear(model_dim, head_count * self.dim_per_head)
self.linear_values = nn.Linear(model_dim, head_count * self.
dim_per_head)
self.linear_query = nn.Linear(model_dim, head_count * self.dim_per_head
)
self.sm = nn.Softmax(dim=-1)
self.dropout = nn.Dropout(dropout)
self.final_linear = nn.Linear(model_dim, model_dim)
def _split_heads(self, x):
batch_size = x.size(0)
return x.view(batch_size, -1, self.head_count, self.dim_per_head
).transpose(1, 2)
def _combine_heads(self, x):
""":param x: [batch_size * head_count, seq_len, dim_per_head]"""
seq_len = x.size(2)
return x.transpose(1, 2).contiguous().view(-1, seq_len, self.
head_count * self.dim_per_head)
def compute_cache(self, key, value):
key_up = self._split_heads(self.linear_keys(key))
value_up = self._split_heads(self.linear_values(value))
return [key_up, value_up]
def forward(self, key, value, query, mask=None, cache=None):
"""
Compute the context vector and the attention vectors.
Args:
key (`FloatTensor`): set of `key_len`
key vectors `[batch, key_len, dim]`
value (`FloatTensor`): set of `key_len`
value vectors `[batch, key_len, dim]`
query (`FloatTensor`): set of `query_len`
query vectors `[batch, query_len, dim]`
mask: binary mask indicating which keys have
non-zero attention `[batch, query_len, key_len]`
Returns:
(`FloatTensor`, `FloatTensor`) :
* output context vectors `[batch, query_len, dim]`
* one of the attention vectors `[batch, query_len, key_len]`
"""
batch_size = key.size(0)
dim_per_head = self.dim_per_head
head_count = self.head_count
key_len = key.size(1)
query_len = query.size(1)
if cache is not None:
key_up, value_up = cache
else:
key_up = self._split_heads(self.linear_keys(key))
value_up = self._split_heads(self.linear_values(value))
query_up = self._split_heads(self.linear_query(query))
query_up = query_up / math.sqrt(dim_per_head)
scores = torch.matmul(query_up, key_up.transpose(2, 3))
if mask is not None:
mask = mask.unsqueeze(1).expand_as(scores)
scores = scores.masked_fill(Variable(mask), -1e+18)
attn = self.sm(scores)
drop_attn = self.dropout(attn)
context = self._combine_heads(torch.matmul(drop_attn, value_up))
output = self.final_linear(context)
top_attn = attn.view(batch_size, head_count, query_len, key_len)[:,
0, :, :].contiguous()
return output, top_attn
class PositionwiseFeedForward(nn.Module):
""" A two-layer Feed-Forward-Network with residual layer norm.
Args:
size (int): the size of input for the first-layer of the FFN.
hidden_size (int): the hidden layer size of the second-layer
of the FNN.
dropout (float): dropout probability(0-1.0).
"""
def __init__(self, size, hidden_size, dropout=0.1):
super(PositionwiseFeedForward, self).__init__()
self.w_1 = nn.Linear(size, hidden_size)
self.w_2 = nn.Linear(hidden_size, size)
self.layer_norm = LayerNorm(size)
self.dropout_1 = nn.Dropout(dropout, inplace=True)
self.relu = nn.ReLU(inplace=True)
self.dropout_2 = nn.Dropout(dropout)
def forward(self, x):
inter = self.dropout_1(self.relu(self.w_1(self.layer_norm(x))))
output = self.dropout_2(self.w_2(inter))
return output + x
class DecoderBlock(nn.Module):
""" Compose with three layers """
def __init__(self, d_model, d_inner_hid, n_head, dropout=0.1):
super(DecoderBlock, self).__init__()
self.slf_attn = MultiHeadedAttention(head_count=n_head, model_dim=
d_model, dropout=dropout)
self.ctx_attn = MultiHeadedAttention(head_count=n_head, model_dim=
d_model, dropout=dropout)
self.pos_ffn = PositionwiseFeedForward(size=d_model, hidden_size=
d_inner_hid)
self.layer_norm_1 = LayerNorm(d_model)
self.layer_norm_2 = LayerNorm(d_model)
self.dropout = nn.Dropout(dropout)
def compute_cache(self, enc_output):
return self.ctx_attn.compute_cache(enc_output, enc_output)
def forward(self, dec_input, enc_output, slf_attn_mask=None,
dec_enc_attn_mask=None, cache=None):
_input_batch, _input_len, _ = dec_input.size()
_contxt_batch, _contxt_len, _ = enc_output.size()
input_norm = self.layer_norm_1(dec_input)
all_input = input_norm
query, _ = self.slf_attn(all_input, all_input, input_norm, mask=
slf_attn_mask)
query = self.dropout(query) + dec_input
query_norm = self.layer_norm_2(query)
mid, attn = self.ctx_attn(enc_output, enc_output, query_norm, mask=
dec_enc_attn_mask, cache=cache)
output = self.pos_ffn(self.dropout(mid) + query)
return output, attn, all_input
def get_inputs():
return [torch.rand([4, 4, 4]), torch.rand([4, 4, 4])]
def get_init_inputs():
return [[], {'d_model': 4, 'd_inner_hid': 4, 'n_head': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
import math
from torch.autograd import Variable
import torch.nn as nn
import torch.optim
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_add_div_mean_mul_std_sub_0(in_ptr0, in_ptr1, in_ptr2,
out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 4
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + x2, xmask)
tmp2 = tl.load(in_ptr1 + 4 * x1, xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr1 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr1 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr1 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp30 = tl.load(in_ptr2 + x0, xmask, eviction_policy='evict_last')
tmp4 = tmp2 + tmp3
tmp6 = tmp4 + tmp5
tmp8 = tmp6 + tmp7
tmp9 = 4.0
tmp10 = tmp8 / tmp9
tmp11 = tmp1 - tmp10
tmp12 = tmp0 * tmp11
tmp13 = tmp2 - tmp10
tmp14 = tmp13 * tmp13
tmp15 = tmp3 - tmp10
tmp16 = tmp15 * tmp15
tmp17 = tmp14 + tmp16
tmp18 = tmp5 - tmp10
tmp19 = tmp18 * tmp18
tmp20 = tmp17 + tmp19
tmp21 = tmp7 - tmp10
tmp22 = tmp21 * tmp21
tmp23 = tmp20 + tmp22
tmp24 = 3.0
tmp25 = tmp23 / tmp24
tmp26 = libdevice.sqrt(tmp25)
tmp27 = 1e-06
tmp28 = tmp26 + tmp27
tmp29 = tmp12 / tmp28
tmp31 = tmp29 + tmp30
tl.store(out_ptr0 + x2, tmp31, xmask)
@triton.jit
def triton_poi_fused_clone_div_1(in_ptr0, in_ptr1, out_ptr0, ynumel, xnumel,
YBLOCK: tl.constexpr, XBLOCK: tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = yindex // 4
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + 4 * x2 + 16 * y1), xmask & ymask,
eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + y0, ymask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = 1.0
tmp4 = tmp2 * tmp3
tl.store(out_ptr0 + (x2 + 4 * y3), tmp4, xmask & ymask)
@triton.jit
def triton_poi_fused_clone_2(in_ptr0, in_ptr1, out_ptr0, ynumel, xnumel,
YBLOCK: tl.constexpr, XBLOCK: tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = yindex // 4
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + 4 * x2 + 16 * y1), xmask & ymask,
eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + y0, ymask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(out_ptr0 + (x2 + 4 * y3), tmp2, xmask & ymask)
@triton.jit
def triton_poi_fused__softmax_3(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp3 = triton_helpers.maximum(tmp1, tmp2)
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp7 = triton_helpers.maximum(tmp5, tmp6)
tmp8 = tmp0 - tmp7
tmp9 = tl_math.exp(tmp8)
tl.store(out_ptr0 + x2, tmp9, xmask)
@triton.jit
def triton_poi_fused__softmax_4(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + x2, tmp8, xmask)
@triton.jit
def triton_poi_fused_clone_5(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK: tl.
constexpr, XBLOCK: tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = yindex // 4
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + 4 * x2 + 16 * y1), xmask & ymask,
eviction_policy='evict_last')
tl.store(out_ptr0 + (x2 + 4 * y3), tmp0, xmask & ymask)
@triton.jit
def triton_poi_fused_add_mean_std_6(in_out_ptr0, in_ptr0, in_ptr1, out_ptr0,
xnumel, XBLOCK: tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + 4 * x0, xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + 4 * x0, xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr0 + (1 + 4 * x0), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr1 + (1 + 4 * x0), xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr0 + (2 + 4 * x0), xmask, eviction_policy='evict_last')
tmp8 = tl.load(in_ptr1 + (2 + 4 * x0), xmask, eviction_policy='evict_last')
tmp11 = tl.load(in_ptr0 + (3 + 4 * x0), xmask, eviction_policy='evict_last'
)
tmp12 = tl.load(in_ptr1 + (3 + 4 * x0), xmask, eviction_policy='evict_last'
)
tmp2 = tmp0 + tmp1
tmp5 = tmp3 + tmp4
tmp6 = tmp2 + tmp5
tmp9 = tmp7 + tmp8
tmp10 = tmp6 + tmp9
tmp13 = tmp11 + tmp12
tmp14 = tmp10 + tmp13
tmp15 = 4.0
tmp16 = tmp14 / tmp15
tmp17 = tmp2 - tmp16
tmp18 = tmp17 * tmp17
tmp19 = tmp5 - tmp16
tmp20 = tmp19 * tmp19
tmp21 = tmp18 + tmp20
tmp22 = tmp9 - tmp16
tmp23 = tmp22 * tmp22
tmp24 = tmp21 + tmp23
tmp25 = tmp13 - tmp16
tmp26 = tmp25 * tmp25
tmp27 = tmp24 + tmp26
tmp28 = 3.0
tmp29 = tmp27 / tmp28
tl.store(in_out_ptr0 + x0, tmp29, xmask)
tl.store(out_ptr0 + x0, tmp16, xmask)
@triton.jit
def triton_poi_fused_add_div_mean_mul_std_sub_7(in_ptr0, in_ptr1, in_ptr2,
in_ptr3, in_ptr4, in_ptr5, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 4
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + x2, xmask)
tmp2 = tl.load(in_ptr2 + x2, xmask)
tmp4 = tl.load(in_ptr3 + x1, xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr4 + x1, xmask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr5 + x0, xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 - tmp4
tmp6 = tmp0 * tmp5
tmp8 = libdevice.sqrt(tmp7)
tmp9 = 1e-06
tmp10 = tmp8 + tmp9
tmp11 = tmp6 / tmp10
tmp13 = tmp11 + tmp12
tl.store(out_ptr0 + x2, tmp13, xmask)
@triton.jit
def triton_poi_fused_clone_8(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 16
x1 = xindex // 16
x2 = xindex
tmp0 = tl.load(in_ptr0 + (x0 + 64 * x1), xmask)
tl.store(out_ptr0 + x2, tmp0, xmask)
@triton.jit
def triton_poi_fused_add_9(in_out_ptr0, in_ptr0, in_ptr1, in_ptr2, xnumel,
XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr1 + x2, xmask)
tmp4 = tl.load(in_ptr2 + x2, xmask)
tmp2 = tmp0 + tmp1
tmp5 = tmp3 + tmp4
tmp6 = tmp2 + tmp5
tl.store(in_out_ptr0 + x2, tmp6, xmask)
@triton.jit
def triton_poi_fused_relu_threshold_backward_10(in_out_ptr0, in_ptr0,
out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(in_out_ptr0 + x2, tmp4, xmask)
tl.store(out_ptr0 + x2, tmp6, xmask)
@triton.jit
def triton_poi_fused_view_11(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 4
x1 = xindex // 4
x2 = xindex
tmp0 = tl.load(in_ptr0 + (x0 + 4 * x1 + 16 * (x1 % 4 // 4)), xmask)
tl.store(out_ptr0 + x2, tmp0, xmask)
@triton.jit
def triton_poi_fused_add_12(in_out_ptr0, in_ptr0, in_ptr1, xnumel, XBLOCK:
tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr1 + x2, xmask)
tmp2 = tmp0 + tmp1
tmp4 = tmp2 + tmp3
tl.store(in_out_ptr0 + x2, tmp4, xmask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7, primals_8, primals_9, primals_10, primals_11, primals_12,
primals_13, primals_14, primals_15, primals_16, primals_17,
primals_18, primals_19, primals_20, primals_21, primals_22,
primals_23, primals_24, primals_25, primals_26, primals_27, primals_28
) = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_2, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_3, (4,), (1,))
assert_size_stride(primals_4, (4,), (1,))
assert_size_stride(primals_5, (4, 4), (4, 1))
assert_size_stride(primals_6, (4,), (1,))
assert_size_stride(primals_7, (4, 4), (4, 1))
assert_size_stride(primals_8, (4,), (1,))
assert_size_stride(primals_9, (4, 4), (4, 1))
assert_size_stride(primals_10, (4,), (1,))
assert_size_stride(primals_11, (4, 4), (4, 1))
assert_size_stride(primals_12, (4,), (1,))
assert_size_stride(primals_13, (4,), (1,))
assert_size_stride(primals_14, (4,), (1,))
assert_size_stride(primals_15, (4, 4), (4, 1))
assert_size_stride(primals_16, (4,), (1,))
assert_size_stride(primals_17, (4, 4), (4, 1))
assert_size_stride(primals_18, (4,), (1,))
assert_size_stride(primals_19, (4, 4), (4, 1))
assert_size_stride(primals_20, (4,), (1,))
assert_size_stride(primals_21, (4, 4), (4, 1))
assert_size_stride(primals_22, (4,), (1,))
assert_size_stride(primals_23, (4,), (1,))
assert_size_stride(primals_24, (4,), (1,))
assert_size_stride(primals_25, (4, 4), (4, 1))
assert_size_stride(primals_26, (4,), (1,))
assert_size_stride(primals_27, (4, 4), (4, 1))
assert_size_stride(primals_28, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_add_div_mean_mul_std_sub_0[grid(64)](primals_3,
primals_1, primals_4, buf0, 64, XBLOCK=64, num_warps=1,
num_stages=1)
del primals_3
del primals_4
buf1 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(buf0, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_5, (4, 4), (1, 4), 0), out=buf1)
buf2 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(buf0, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_7, (4, 4), (1, 4), 0), out=buf2)
buf3 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(buf0, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_9, (4, 4), (1, 4), 0), out=buf3)
buf4 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 1), torch.float32)
triton_poi_fused_clone_div_1[grid(16, 4)](buf3, primals_10, buf4,
16, 4, XBLOCK=4, YBLOCK=8, num_warps=1, num_stages=1)
del primals_10
buf5 = reinterpret_tensor(buf3, (4, 4, 1, 4), (16, 4, 4, 1), 0)
del buf3
triton_poi_fused_clone_2[grid(16, 4)](buf1, primals_6, buf5, 16, 4,
XBLOCK=4, YBLOCK=8, num_warps=1, num_stages=1)
del primals_6
buf6 = empty_strided_cuda((16, 4, 4), (16, 4, 1), torch.float32)
extern_kernels.bmm(reinterpret_tensor(buf4, (16, 4, 1), (4, 1, 0),
0), reinterpret_tensor(buf5, (16, 1, 4), (4, 0, 1), 0), out=buf6)
buf7 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
triton_poi_fused__softmax_3[grid(256)](buf6, buf7, 256, XBLOCK=128,
num_warps=4, num_stages=1)
buf8 = reinterpret_tensor(buf6, (4, 4, 4, 4), (64, 16, 4, 1), 0)
del buf6
triton_poi_fused__softmax_4[grid(256)](buf7, buf8, 256, XBLOCK=128,
num_warps=4, num_stages=1)
buf9 = reinterpret_tensor(buf1, (4, 4, 4, 1), (16, 4, 1, 1), 0)
del buf1
triton_poi_fused_clone_2[grid(16, 4)](buf2, primals_8, buf9, 16, 4,
XBLOCK=4, YBLOCK=8, num_warps=1, num_stages=1)
del primals_8
buf10 = reinterpret_tensor(buf2, (16, 4, 1), (4, 1, 1), 0)
del buf2
extern_kernels.bmm(reinterpret_tensor(buf8, (16, 4, 4), (16, 4, 1),
0), reinterpret_tensor(buf9, (16, 4, 1), (4, 1, 0), 0), out=buf10)
buf11 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 1), torch.float32)
triton_poi_fused_clone_5[grid(16, 4)](buf10, buf11, 16, 4, XBLOCK=4,
YBLOCK=16, num_warps=1, num_stages=1)
buf12 = reinterpret_tensor(buf10, (16, 4), (4, 1), 0)
del buf10
extern_kernels.addmm(primals_12, reinterpret_tensor(buf11, (16, 4),
(4, 1), 0), reinterpret_tensor(primals_11, (4, 4), (1, 4), 0),
alpha=1, beta=1, out=buf12)
del primals_12
buf13 = empty_strided_cuda((4, 4, 1), (4, 1, 16), torch.float32)
buf14 = buf13
del buf13
buf15 = empty_strided_cuda((4, 4, 1), (4, 1, 16), torch.float32)
triton_poi_fused_add_mean_std_6[grid(16)](buf14, buf12, primals_1,
buf15, 16, XBLOCK=16, num_warps=1, num_stages=1)
buf16 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_2, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_15, (4, 4), (1, 4), 0), out=buf16)
del primals_15
buf17 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_2, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_17, (4, 4), (1, 4), 0), out=buf17)
del primals_17
buf18 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
triton_poi_fused_add_div_mean_mul_std_sub_7[grid(64)](primals_13,
buf12, primals_1, buf15, buf14, primals_14, buf18, 64, XBLOCK=
64, num_warps=1, num_stages=1)
del buf14
del buf15
del primals_14
buf19 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(buf18, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_19, (4, 4), (1, 4), 0), out=buf19)
buf20 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 1), torch.float32)
triton_poi_fused_clone_div_1[grid(16, 4)](buf19, primals_20, buf20,
16, 4, XBLOCK=4, YBLOCK=8, num_warps=1, num_stages=1)
del primals_20
buf21 = reinterpret_tensor(buf19, (4, 4, 1, 4), (16, 4, 4, 1), 0)
del buf19
triton_poi_fused_clone_2[grid(16, 4)](buf16, primals_16, buf21, 16,
4, XBLOCK=4, YBLOCK=8, num_warps=1, num_stages=1)
del primals_16
buf22 = reinterpret_tensor(buf7, (16, 4, 4), (16, 4, 1), 0)
del buf7
extern_kernels.bmm(reinterpret_tensor(buf20, (16, 4, 1), (4, 1, 0),
0), reinterpret_tensor(buf21, (16, 1, 4), (4, 0, 1), 0), out=buf22)
buf23 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
triton_poi_fused__softmax_3[grid(256)](buf22, buf23, 256, XBLOCK=
128, num_warps=4, num_stages=1)
buf24 = reinterpret_tensor(buf22, (4, 4, 4, 4), (64, 16, 4, 1), 0)
del buf22
triton_poi_fused__softmax_4[grid(256)](buf23, buf24, 256, XBLOCK=
128, num_warps=4, num_stages=1)
del buf23
buf25 = reinterpret_tensor(buf16, (4, 4, 4, 1), (16, 4, 1, 1), 0)
del buf16
triton_poi_fused_clone_2[grid(16, 4)](buf17, primals_18, buf25, 16,
4, XBLOCK=4, YBLOCK=8, num_warps=1, num_stages=1)
del primals_18
buf26 = reinterpret_tensor(buf17, (16, 4, 1), (4, 1, 1), 0)
del buf17
extern_kernels.bmm(reinterpret_tensor(buf24, (16, 4, 4), (16, 4, 1),
0), reinterpret_tensor(buf25, (16, 4, 1), (4, 1, 0), 0), out=buf26)
buf27 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 1), torch.float32)
triton_poi_fused_clone_5[grid(16, 4)](buf26, buf27, 16, 4, XBLOCK=4,
YBLOCK=16, num_warps=1, num_stages=1)
buf28 = reinterpret_tensor(buf26, (16, 4), (4, 1), 0)
del buf26
extern_kernels.mm(reinterpret_tensor(buf27, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_21, (4, 4), (1, 4), 0), out=buf28)
buf29 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
triton_poi_fused_clone_8[grid(64)](buf24, buf29, 64, XBLOCK=64,
num_warps=1, num_stages=1)
buf30 = reinterpret_tensor(buf28, (4, 4, 4), (16, 4, 1), 0)
del buf28
triton_poi_fused_add_9[grid(64)](buf30, primals_22, buf12,
primals_1, 64, XBLOCK=64, num_warps=1, num_stages=1)
del primals_22
buf31 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
triton_poi_fused_add_div_mean_mul_std_sub_0[grid(64)](primals_23,
buf30, primals_24, buf31, 64, XBLOCK=64, num_warps=1, num_stages=1)
del primals_24
buf32 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(buf31, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_25, (4, 4), (1, 4), 0), out=buf32)
buf33 = reinterpret_tensor(buf32, (4, 4, 4), (16, 4, 1), 0)
del buf32
buf37 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.bool)
triton_poi_fused_relu_threshold_backward_10[grid(64)](buf33,
primals_26, buf37, 64, XBLOCK=64, num_warps=1, num_stages=1)
del primals_26
buf34 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
triton_poi_fused_view_11[grid(64)](buf33, buf34, 64, XBLOCK=64,
num_warps=1, num_stages=1)
buf35 = reinterpret_tensor(buf33, (16, 4), (4, 1), 0)
del buf33
extern_kernels.mm(buf34, reinterpret_tensor(primals_27, (4, 4), (1,
4), 0), out=buf35)
buf36 = reinterpret_tensor(buf35, (4, 4, 4), (16, 4, 1), 0)
del buf35
triton_poi_fused_add_12[grid(64)](buf36, primals_28, buf30, 64,
XBLOCK=64, num_warps=1, num_stages=1)
del primals_28
return (buf36, buf29, buf0, primals_1, primals_13, primals_23, buf0,
buf8, reinterpret_tensor(buf11, (16, 4), (4, 1), 0), buf12,
reinterpret_tensor(primals_2, (16, 4), (4, 1), 0),
reinterpret_tensor(buf18, (16, 4), (4, 1), 0), buf24,
reinterpret_tensor(buf27, (16, 4), (4, 1), 0), buf30,
reinterpret_tensor(buf31, (16, 4), (4, 1), 0), buf34, primals_27,
buf37, primals_25, primals_21, reinterpret_tensor(buf25, (16, 1, 4),
(4, 1, 1), 0), reinterpret_tensor(buf20, (16, 1, 4), (4, 1, 1), 0),
reinterpret_tensor(buf21, (16, 4, 1), (4, 1, 4), 0), primals_19,
primals_11, reinterpret_tensor(buf9, (16, 1, 4), (4, 1, 1), 0),
reinterpret_tensor(buf4, (16, 1, 4), (4, 1, 1), 0),
reinterpret_tensor(buf5, (16, 4, 1), (4, 1, 4), 0), primals_9,
primals_7, primals_5)
class LayerNorm(nn.Module):
def __init__(self, features, eps=1e-06):
super(LayerNorm, self).__init__()
self.a_2 = nn.Parameter(torch.ones(features))
self.b_2 = nn.Parameter(torch.zeros(features))
self.eps = eps
def forward(self, x):
mean = x.mean(-1, keepdim=True)
std = x.std(-1, keepdim=True)
return self.a_2 * (x - mean) / (std + self.eps) + self.b_2
class MultiHeadedAttention(nn.Module):
"""
Multi-Head Attention module from
"Attention is All You Need"
:cite:`DBLP:journals/corr/VaswaniSPUJGKP17`.
Similar to standard `dot` attention but uses
multiple attention distributions simulataneously
to select relevant items.
.. mermaid::
graph BT
A[key]
B[value]
C[query]
O[output]
subgraph Attn
D[Attn 1]
E[Attn 2]
F[Attn N]
end
A --> D
C --> D
A --> E
C --> E
A --> F
C --> F
D --> O
E --> O
F --> O
B --> O
Also includes several additional tricks.
Args:
head_count (int): number of parallel heads
model_dim (int): the dimension of keys/values/queries,
must be divisible by head_count
dropout (float): dropout parameter
"""
def __init__(self, head_count, model_dim, dropout=0.1):
assert model_dim % head_count == 0
self.dim_per_head = model_dim // head_count
self.model_dim = model_dim
super(MultiHeadedAttention, self).__init__()
self.head_count = head_count
self.linear_keys = nn.Linear(model_dim, head_count * self.dim_per_head)
self.linear_values = nn.Linear(model_dim, head_count * self.
dim_per_head)
self.linear_query = nn.Linear(model_dim, head_count * self.dim_per_head
)
self.sm = nn.Softmax(dim=-1)
self.dropout = nn.Dropout(dropout)
self.final_linear = nn.Linear(model_dim, model_dim)
def _split_heads(self, x):
batch_size = x.size(0)
return x.view(batch_size, -1, self.head_count, self.dim_per_head
).transpose(1, 2)
def _combine_heads(self, x):
""":param x: [batch_size * head_count, seq_len, dim_per_head]"""
seq_len = x.size(2)
return x.transpose(1, 2).contiguous().view(-1, seq_len, self.
head_count * self.dim_per_head)
def compute_cache(self, key, value):
key_up = self._split_heads(self.linear_keys(key))
value_up = self._split_heads(self.linear_values(value))
return [key_up, value_up]
def forward(self, key, value, query, mask=None, cache=None):
"""
Compute the context vector and the attention vectors.
Args:
key (`FloatTensor`): set of `key_len`
key vectors `[batch, key_len, dim]`
value (`FloatTensor`): set of `key_len`
value vectors `[batch, key_len, dim]`
query (`FloatTensor`): set of `query_len`
query vectors `[batch, query_len, dim]`
mask: binary mask indicating which keys have
non-zero attention `[batch, query_len, key_len]`
Returns:
(`FloatTensor`, `FloatTensor`) :
* output context vectors `[batch, query_len, dim]`
* one of the attention vectors `[batch, query_len, key_len]`
"""
batch_size = key.size(0)
dim_per_head = self.dim_per_head
head_count = self.head_count
key_len = key.size(1)
query_len = query.size(1)
if cache is not None:
key_up, value_up = cache
else:
key_up = self._split_heads(self.linear_keys(key))
value_up = self._split_heads(self.linear_values(value))
query_up = self._split_heads(self.linear_query(query))
query_up = query_up / math.sqrt(dim_per_head)
scores = torch.matmul(query_up, key_up.transpose(2, 3))
if mask is not None:
mask = mask.unsqueeze(1).expand_as(scores)
scores = scores.masked_fill(Variable(mask), -1e+18)
attn = self.sm(scores)
drop_attn = self.dropout(attn)
context = self._combine_heads(torch.matmul(drop_attn, value_up))
output = self.final_linear(context)
top_attn = attn.view(batch_size, head_count, query_len, key_len)[:,
0, :, :].contiguous()
return output, top_attn
class PositionwiseFeedForward(nn.Module):
""" A two-layer Feed-Forward-Network with residual layer norm.
Args:
size (int): the size of input for the first-layer of the FFN.
hidden_size (int): the hidden layer size of the second-layer
of the FNN.
dropout (float): dropout probability(0-1.0).
"""
def __init__(self, size, hidden_size, dropout=0.1):
super(PositionwiseFeedForward, self).__init__()
self.w_1 = nn.Linear(size, hidden_size)
self.w_2 = nn.Linear(hidden_size, size)
self.layer_norm = LayerNorm(size)
self.dropout_1 = nn.Dropout(dropout, inplace=True)
self.relu = nn.ReLU(inplace=True)
self.dropout_2 = nn.Dropout(dropout)
def forward(self, x):
inter = self.dropout_1(self.relu(self.w_1(self.layer_norm(x))))
output = self.dropout_2(self.w_2(inter))
return output + x
class DecoderBlockNew(nn.Module):
""" Compose with three layers """
def __init__(self, d_model, d_inner_hid, n_head, dropout=0.1):
super(DecoderBlockNew, self).__init__()
self.slf_attn = MultiHeadedAttention(head_count=n_head, model_dim=
d_model, dropout=dropout)
self.ctx_attn = MultiHeadedAttention(head_count=n_head, model_dim=
d_model, dropout=dropout)
self.pos_ffn = PositionwiseFeedForward(size=d_model, hidden_size=
d_inner_hid)
self.layer_norm_1 = LayerNorm(d_model)
self.layer_norm_2 = LayerNorm(d_model)
self.dropout = nn.Dropout(dropout)
def compute_cache(self, enc_output):
return self.ctx_attn.compute_cache(enc_output, enc_output)
def forward(self, input_0, input_1):
primals_5 = self.slf_attn.linear_keys.weight
primals_3 = self.slf_attn.linear_keys.bias
primals_7 = self.slf_attn.linear_values.weight
primals_4 = self.slf_attn.linear_values.bias
primals_9 = self.slf_attn.linear_query.weight
primals_6 = self.slf_attn.linear_query.bias
primals_11 = self.slf_attn.final_linear.weight
primals_8 = self.slf_attn.final_linear.bias
primals_15 = self.ctx_attn.linear_keys.weight
primals_10 = self.ctx_attn.linear_keys.bias
primals_17 = self.ctx_attn.linear_values.weight
primals_12 = self.ctx_attn.linear_values.bias
primals_19 = self.ctx_attn.linear_query.weight
primals_13 = self.ctx_attn.linear_query.bias
primals_21 = self.ctx_attn.final_linear.weight
primals_14 = self.ctx_attn.final_linear.bias
primals_25 = self.pos_ffn.w_1.weight
primals_16 = self.pos_ffn.w_1.bias
primals_27 = self.pos_ffn.w_2.weight
primals_18 = self.pos_ffn.w_2.bias
primals_20 = self.pos_ffn.layer_norm.a_2
primals_22 = self.pos_ffn.layer_norm.b_2
primals_23 = self.layer_norm_1.a_2
primals_24 = self.layer_norm_1.b_2
primals_26 = self.layer_norm_2.a_2
primals_28 = self.layer_norm_2.b_2
primals_1 = input_0
primals_2 = input_1
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7, primals_8, primals_9,
primals_10, primals_11, primals_12, primals_13, primals_14,
primals_15, primals_16, primals_17, primals_18, primals_19,
primals_20, primals_21, primals_22, primals_23, primals_24,
primals_25, primals_26, primals_27, primals_28])
return output[0], output[1], output[2]
| howardchenhd/Transformer-pytorch | DecoderBlock | false | 6,850 | [
"MIT"
] | 1 | ae71ed5767272feb7e717be6d5bfce46f80ec57a | https://github.com/howardchenhd/Transformer-pytorch/tree/ae71ed5767272feb7e717be6d5bfce46f80ec57a | import math
import torch
from torch.autograd import Variable
import torch.nn as nn
import torch.optim
class LayerNorm(nn.Module):
def __init__(self, features, eps=1e-06):
super().__init__()
self.a_2 = nn.Parameter(torch.ones(features))
self.b_2 = nn.Parameter(torch.zeros(features))
self.eps = eps
def forward(self, x):
mean = x.mean(-1, keepdim=True)
std = x.std(-1, keepdim=True)
return self.a_2 * (x - mean) / (std + self.eps) + self.b_2
class MultiHeadedAttention(nn.Module):
"""
Multi-Head Attention module from
"Attention is All You Need"
:cite:`DBLP:journals/corr/VaswaniSPUJGKP17`.
Similar to standard `dot` attention but uses
multiple attention distributions simulataneously
to select relevant items.
.. mermaid::
graph BT
A[key]
B[value]
C[query]
O[output]
subgraph Attn
D[Attn 1]
E[Attn 2]
F[Attn N]
end
A --> D
C --> D
A --> E
C --> E
A --> F
C --> F
D --> O
E --> O
F --> O
B --> O
Also includes several additional tricks.
Args:
head_count (int): number of parallel heads
model_dim (int): the dimension of keys/values/queries,
must be divisible by head_count
dropout (float): dropout parameter
"""
def __init__(self, head_count, model_dim, dropout=0.1):
assert model_dim % head_count == 0
self.dim_per_head = model_dim // head_count
self.model_dim = model_dim
super().__init__()
self.head_count = head_count
self.linear_keys = nn.Linear(model_dim, head_count * self.dim_per_head)
self.linear_values = nn.Linear(model_dim, head_count * self.
dim_per_head)
self.linear_query = nn.Linear(model_dim, head_count * self.dim_per_head
)
self.sm = nn.Softmax(dim=-1)
self.dropout = nn.Dropout(dropout)
self.final_linear = nn.Linear(model_dim, model_dim)
def _split_heads(self, x):
batch_size = x.size(0)
return x.view(batch_size, -1, self.head_count, self.dim_per_head
).transpose(1, 2)
def _combine_heads(self, x):
""":param x: [batch_size * head_count, seq_len, dim_per_head]"""
seq_len = x.size(2)
return x.transpose(1, 2).contiguous().view(-1, seq_len, self.
head_count * self.dim_per_head)
def compute_cache(self, key, value):
key_up = self._split_heads(self.linear_keys(key))
value_up = self._split_heads(self.linear_values(value))
return [key_up, value_up]
def forward(self, key, value, query, mask=None, cache=None):
"""
Compute the context vector and the attention vectors.
Args:
key (`FloatTensor`): set of `key_len`
key vectors `[batch, key_len, dim]`
value (`FloatTensor`): set of `key_len`
value vectors `[batch, key_len, dim]`
query (`FloatTensor`): set of `query_len`
query vectors `[batch, query_len, dim]`
mask: binary mask indicating which keys have
non-zero attention `[batch, query_len, key_len]`
Returns:
(`FloatTensor`, `FloatTensor`) :
* output context vectors `[batch, query_len, dim]`
* one of the attention vectors `[batch, query_len, key_len]`
"""
batch_size = key.size(0)
dim_per_head = self.dim_per_head
head_count = self.head_count
key_len = key.size(1)
query_len = query.size(1)
if cache is not None:
key_up, value_up = cache
else:
key_up = self._split_heads(self.linear_keys(key))
value_up = self._split_heads(self.linear_values(value))
query_up = self._split_heads(self.linear_query(query))
# ... truncated (>4000 chars) for memory efficiency |
Biaffine | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/am/cambydvhijs7vhod3hwt5aje7cqigviqetkgm6iccyrgbwhmprcb.py
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.cat]
# Source node to ATen node mapping:
# x => cat
# Graph fragment:
# %cat : [num_users=1] = call_function[target=torch.ops.aten.cat.default](args = ([%primals_1, %full_default], -1), kwargs = {})
triton_poi_fused_cat_0 = async_compile.triton('triton_poi_fused_cat_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[512],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_cat_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_cat_0(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 320
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 5
x1 = (xindex // 5)
x2 = xindex
tmp0 = x0
tmp1 = tl.full([1], 0, tl.int64)
tmp2 = tmp0 >= tmp1
tmp3 = tl.full([1], 4, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = tl.load(in_ptr0 + ((4*x1) + x0), tmp4 & xmask, eviction_policy='evict_last', other=0.0)
tmp6 = tmp0 >= tmp3
tmp7 = tl.full([1], 5, tl.int64)
tmp8 = tmp0 < tmp7
tmp9 = 1.0
tmp10 = tl.full(tmp9.shape, 0.0, tmp9.dtype)
tmp11 = tl.where(tmp6, tmp9, tmp10)
tmp12 = tl.where(tmp4, tmp5, tmp11)
tl.store(out_ptr0 + (x2), tmp12, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_3, (1, 5, 5), (25, 5, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4, 5), (80, 20, 5, 1), torch.float32)
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.cat]
stream0 = get_raw_stream(0)
triton_poi_fused_cat_0.run(primals_1, buf0, 320, grid=grid(320), stream=stream0)
del primals_1
buf1 = empty_strided_cuda((16, 4, 5), (20, 5, 1), torch.float32)
# Topologically Sorted Source Nodes: [matmul], Original ATen: [aten.bmm]
extern_kernels.bmm(reinterpret_tensor(buf0, (16, 4, 5), (20, 5, 1), 0), reinterpret_tensor(primals_3, (16, 5, 5), (0, 5, 1), 0), out=buf1)
del primals_3
buf2 = empty_strided_cuda((4, 4, 4, 5), (80, 20, 5, 1), torch.float32)
# Topologically Sorted Source Nodes: [y], Original ATen: [aten.cat]
triton_poi_fused_cat_0.run(primals_2, buf2, 320, grid=grid(320), stream=stream0)
del primals_2
buf3 = empty_strided_cuda((16, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [s], Original ATen: [aten.bmm]
extern_kernels.bmm(buf1, reinterpret_tensor(buf2, (16, 5, 4), (20, 1, 5), 0), out=buf3)
del buf1
return (reinterpret_tensor(buf3, (4, 4, 4, 4), (64, 16, 4, 1), 0), reinterpret_tensor(buf2, (16, 4, 5), (20, 5, 1), 0), reinterpret_tensor(buf0, (16, 5, 4), (20, 1, 5), 0), )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((1, 5, 5), (25, 5, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
class Biaffine(nn.Module):
def __init__(self, n_in, n_out=1, bias_x=True, bias_y=True):
super(Biaffine, self).__init__()
self.n_in = n_in
self.n_out = n_out
self.bias_x = bias_x
self.bias_y = bias_y
self.weight = nn.Parameter(torch.Tensor(n_out, n_in + bias_x, n_in +
bias_y))
self.reset_parameters()
def extra_repr(self):
info = f'n_in={self.n_in}, n_out={self.n_out}'
if self.bias_x:
info += f', bias_x={self.bias_x}'
if self.bias_y:
info += f', bias_y={self.bias_y}'
return info
def reset_parameters(self):
nn.init.zeros_(self.weight)
def forward(self, x, y):
if self.bias_x:
x = torch.cat([x, x.new_ones(x.shape[:-1]).unsqueeze(-1)], -1)
if self.bias_y:
y = torch.cat([y, y.new_ones(y.shape[:-1]).unsqueeze(-1)], -1)
x = x.unsqueeze(1)
y = y.unsqueeze(1)
s = x @ self.weight @ torch.transpose(y, -1, -2)
s = s.squeeze(1)
return s
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'n_in': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_cat_0(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 320
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 5
x1 = xindex // 5
x2 = xindex
tmp0 = x0
tl.full([1], 0, tl.int64)
tmp3 = tl.full([1], 4, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = tl.load(in_ptr0 + (4 * x1 + x0), tmp4 & xmask, eviction_policy=
'evict_last', other=0.0)
tmp6 = tmp0 >= tmp3
tl.full([1], 5, tl.int64)
tmp9 = 1.0
tmp10 = tl.full(tmp9.shape, 0.0, tmp9.dtype)
tmp11 = tl.where(tmp6, tmp9, tmp10)
tmp12 = tl.where(tmp4, tmp5, tmp11)
tl.store(out_ptr0 + x2, tmp12, xmask)
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_3, (1, 5, 5), (25, 5, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4, 5), (80, 20, 5, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_cat_0[grid(320)](primals_1, buf0, 320, XBLOCK=256,
num_warps=4, num_stages=1)
del primals_1
buf1 = empty_strided_cuda((16, 4, 5), (20, 5, 1), torch.float32)
extern_kernels.bmm(reinterpret_tensor(buf0, (16, 4, 5), (20, 5, 1),
0), reinterpret_tensor(primals_3, (16, 5, 5), (0, 5, 1), 0),
out=buf1)
del primals_3
buf2 = empty_strided_cuda((4, 4, 4, 5), (80, 20, 5, 1), torch.float32)
triton_poi_fused_cat_0[grid(320)](primals_2, buf2, 320, XBLOCK=256,
num_warps=4, num_stages=1)
del primals_2
buf3 = empty_strided_cuda((16, 4, 4), (16, 4, 1), torch.float32)
extern_kernels.bmm(buf1, reinterpret_tensor(buf2, (16, 5, 4), (20,
1, 5), 0), out=buf3)
del buf1
return reinterpret_tensor(buf3, (4, 4, 4, 4), (64, 16, 4, 1), 0
), reinterpret_tensor(buf2, (16, 4, 5), (20, 5, 1), 0
), reinterpret_tensor(buf0, (16, 5, 4), (20, 1, 5), 0)
class BiaffineNew(nn.Module):
def __init__(self, n_in, n_out=1, bias_x=True, bias_y=True):
super(BiaffineNew, self).__init__()
self.n_in = n_in
self.n_out = n_out
self.bias_x = bias_x
self.bias_y = bias_y
self.weight = nn.Parameter(torch.Tensor(n_out, n_in + bias_x, n_in +
bias_y))
self.reset_parameters()
def extra_repr(self):
info = f'n_in={self.n_in}, n_out={self.n_out}'
if self.bias_x:
info += f', bias_x={self.bias_x}'
if self.bias_y:
info += f', bias_y={self.bias_y}'
return info
def reset_parameters(self):
nn.init.zeros_(self.weight)
def forward(self, input_0, input_1):
primals_3 = self.weight
primals_1 = input_0
primals_2 = input_1
output = call([primals_1, primals_2, primals_3])
return output[0]
| icewing1996/biaffine-parser | Biaffine | false | 6,851 | [
"MIT"
] | 1 | f5a4ece7ba9a087d81b76dd6a8ea6aa7d90c6c82 | https://github.com/icewing1996/biaffine-parser/tree/f5a4ece7ba9a087d81b76dd6a8ea6aa7d90c6c82 | import torch
import torch.nn as nn
class Model(nn.Module):
def __init__(self, n_in, n_out=1, bias_x=True, bias_y=True):
super().__init__()
self.n_in = n_in
self.n_out = n_out
self.bias_x = bias_x
self.bias_y = bias_y
self.weight = nn.Parameter(torch.Tensor(n_out, n_in + bias_x, n_in +
bias_y))
self.reset_parameters()
def extra_repr(self):
info = f'n_in={self.n_in}, n_out={self.n_out}'
if self.bias_x:
info += f', bias_x={self.bias_x}'
if self.bias_y:
info += f', bias_y={self.bias_y}'
return info
def reset_parameters(self):
nn.init.zeros_(self.weight)
def forward(self, x, y):
if self.bias_x:
x = torch.cat([x, x.new_ones(x.shape[:-1]).unsqueeze(-1)], -1)
if self.bias_y:
y = torch.cat([y, y.new_ones(y.shape[:-1]).unsqueeze(-1)], -1)
x = x.unsqueeze(1)
y = y.unsqueeze(1)
s = x @ self.weight @ torch.transpose(y, -1, -2)
s = s.squeeze(1)
return s
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4]
|
Critic | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/ms/cmsuzohbg5nq52jnvirovzkvykrzzko5xomu7zyu5e5u2lhegppw.py
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.cat]
# Source node to ATen node mapping:
# x => cat
# Graph fragment:
# %cat : [num_users=2] = call_function[target=torch.ops.aten.cat.default](args = ([%primals_1, %primals_2], 1), kwargs = {})
triton_poi_fused_cat_0 = async_compile.triton('triton_poi_fused_cat_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[32],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_cat_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_cat_0(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 32
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 8
x1 = (xindex // 8)
x2 = xindex
tmp0 = x0
tmp1 = tl.full([1], 0, tl.int64)
tmp2 = tmp0 >= tmp1
tmp3 = tl.full([1], 4, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = tl.load(in_ptr0 + ((4*x1) + x0), tmp4 & xmask, eviction_policy='evict_last', other=0.0)
tmp6 = tmp0 >= tmp3
tmp7 = tl.full([1], 8, tl.int64)
tmp8 = tmp0 < tmp7
tmp9 = tl.load(in_ptr1 + ((4*x1) + ((-4) + x0)), tmp6 & xmask, eviction_policy='evict_last', other=0.0)
tmp10 = tl.where(tmp4, tmp5, tmp9)
tl.store(out_ptr0 + (x2), tmp10, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/c2/cc2i6ob54hxywcta6oeintp4auwrjofyaaspcy67zuwz4bgebehb.py
# Topologically Sorted Source Nodes: [x_2], Original ATen: [aten.native_layer_norm]
# Source node to ATen node mapping:
# x_2 => add, rsqrt, var_mean
# Graph fragment:
# %var_mean : [num_users=2] = call_function[target=torch.ops.aten.var_mean.correction](args = (%addmm, [1]), kwargs = {correction: 0, keepdim: True})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%getitem, 1e-05), kwargs = {})
# %rsqrt : [num_users=1] = call_function[target=torch.ops.aten.rsqrt.default](args = (%add,), kwargs = {})
triton_poi_fused_native_layer_norm_1 = async_compile.triton('triton_poi_fused_native_layer_norm_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[4],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_native_layer_norm_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 4, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_native_layer_norm_1(in_ptr0, out_ptr0, out_ptr1, xnumel, XBLOCK : tl.constexpr):
xnumel = 4
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (4*x0), xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr0 + (1 + (4*x0)), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr0 + (2 + (4*x0)), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (3 + (4*x0)), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp4 = tmp2 + tmp3
tmp6 = tmp4 + tmp5
tmp7 = 4.0
tmp8 = tmp6 / tmp7
tmp9 = tmp0 - tmp8
tmp10 = tmp9 * tmp9
tmp11 = tmp1 - tmp8
tmp12 = tmp11 * tmp11
tmp13 = tmp10 + tmp12
tmp14 = tmp3 - tmp8
tmp15 = tmp14 * tmp14
tmp16 = tmp13 + tmp15
tmp17 = tmp5 - tmp8
tmp18 = tmp17 * tmp17
tmp19 = tmp16 + tmp18
tmp20 = tmp19 / tmp7
tmp21 = 1e-05
tmp22 = tmp20 + tmp21
tmp23 = libdevice.rsqrt(tmp22)
tl.store(out_ptr0 + (x0), tmp8, xmask)
tl.store(out_ptr1 + (x0), tmp23, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/xu/cxugypfyur5ricnklmp5xmosygv56nhapaux5iz63gnadqinxxjs.py
# Topologically Sorted Source Nodes: [x_2, x_3], Original ATen: [aten.native_layer_norm, aten.tanh]
# Source node to ATen node mapping:
# x_2 => add, add_1, mul, mul_1, rsqrt, sub, var_mean
# x_3 => tanh
# Graph fragment:
# %var_mean : [num_users=2] = call_function[target=torch.ops.aten.var_mean.correction](args = (%addmm, [1]), kwargs = {correction: 0, keepdim: True})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%getitem, 1e-05), kwargs = {})
# %rsqrt : [num_users=1] = call_function[target=torch.ops.aten.rsqrt.default](args = (%add,), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%addmm, %getitem_1), kwargs = {})
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sub, %rsqrt), kwargs = {})
# %mul_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%mul, %primals_5), kwargs = {})
# %add_1 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul_1, %primals_6), kwargs = {})
# %tanh : [num_users=2] = call_function[target=torch.ops.aten.tanh.default](args = (%add_1,), kwargs = {})
triton_poi_fused_native_layer_norm_tanh_2 = async_compile.triton('triton_poi_fused_native_layer_norm_tanh_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: '*fp32', 5: '*fp32', 6: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5, 6), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_native_layer_norm_tanh_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_native_layer_norm_tanh_2(in_ptr0, in_ptr1, in_ptr2, in_ptr3, in_ptr4, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
x0 = xindex % 4
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr1 + (x1), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr2 + (x1), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr3 + (x0), xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr4 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 - tmp1
tmp4 = tmp2 * tmp3
tmp6 = tmp4 * tmp5
tmp8 = tmp6 + tmp7
tmp9 = libdevice.tanh(tmp8)
tl.store(out_ptr0 + (x2), tmp9, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9, primals_10, primals_11, primals_12 = args
args.clear()
assert_size_stride(primals_1, (4, 4), (4, 1))
assert_size_stride(primals_2, (4, 4), (4, 1))
assert_size_stride(primals_3, (4, 8), (8, 1))
assert_size_stride(primals_4, (4, ), (1, ))
assert_size_stride(primals_5, (4, ), (1, ))
assert_size_stride(primals_6, (4, ), (1, ))
assert_size_stride(primals_7, (4, 4), (4, 1))
assert_size_stride(primals_8, (4, ), (1, ))
assert_size_stride(primals_9, (4, ), (1, ))
assert_size_stride(primals_10, (4, ), (1, ))
assert_size_stride(primals_11, (1, 4), (4, 1))
assert_size_stride(primals_12, (1, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 8), (8, 1), torch.float32)
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.cat]
stream0 = get_raw_stream(0)
triton_poi_fused_cat_0.run(primals_1, primals_2, buf0, 32, grid=grid(32), stream=stream0)
del primals_1
del primals_2
buf1 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [x_1], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_4, buf0, reinterpret_tensor(primals_3, (8, 4), (1, 8), 0), alpha=1, beta=1, out=buf1)
del primals_3
del primals_4
buf2 = empty_strided_cuda((4, 1), (1, 4), torch.float32)
buf3 = empty_strided_cuda((4, 1), (1, 4), torch.float32)
# Topologically Sorted Source Nodes: [x_2], Original ATen: [aten.native_layer_norm]
triton_poi_fused_native_layer_norm_1.run(buf1, buf2, buf3, 4, grid=grid(4), stream=stream0)
buf4 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [x_2, x_3], Original ATen: [aten.native_layer_norm, aten.tanh]
triton_poi_fused_native_layer_norm_tanh_2.run(buf1, buf2, buf3, primals_5, primals_6, buf4, 16, grid=grid(16), stream=stream0)
del primals_6
buf5 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [x_4], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_8, buf4, reinterpret_tensor(primals_7, (4, 4), (1, 4), 0), alpha=1, beta=1, out=buf5)
del primals_8
buf6 = buf3; del buf3 # reuse
buf7 = buf2; del buf2 # reuse
# Topologically Sorted Source Nodes: [x_5], Original ATen: [aten.native_layer_norm]
triton_poi_fused_native_layer_norm_1.run(buf5, buf6, buf7, 4, grid=grid(4), stream=stream0)
buf8 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [x_5, x_6], Original ATen: [aten.native_layer_norm, aten.tanh]
triton_poi_fused_native_layer_norm_tanh_2.run(buf5, buf6, buf7, primals_9, primals_10, buf8, 16, grid=grid(16), stream=stream0)
del buf6
del primals_10
buf10 = reinterpret_tensor(buf7, (4, 1), (1, 1), 0); del buf7 # reuse
# Topologically Sorted Source Nodes: [V], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_12, buf8, reinterpret_tensor(primals_11, (4, 1), (1, 4), 0), alpha=1, beta=1, out=buf10)
del primals_12
return (buf10, primals_5, primals_9, buf0, buf1, buf4, buf5, buf8, primals_11, primals_7, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 8), (8, 1), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_8 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_9 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_10 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_11 = rand_strided((1, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_12 = rand_strided((1, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9, primals_10, primals_11, primals_12])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn.functional as F
import torch.nn as nn
class Critic(nn.Module):
def __init__(self, hidden_size, num_inputs, action_space):
super(Critic, self).__init__()
self.action_space = action_space
num_outputs = action_space.shape[0]
self.linear1 = nn.Linear(num_inputs + num_outputs, hidden_size)
nn.init.normal_(self.linear1.weight, 0.0, 0.02)
self.ln1 = nn.LayerNorm(hidden_size)
self.linear2 = nn.Linear(hidden_size, hidden_size)
nn.init.normal_(self.linear2.weight, 0.0, 0.02)
self.ln2 = nn.LayerNorm(hidden_size)
self.V = nn.Linear(hidden_size, 1)
torch.nn.init.uniform_(self.V.weight, a=-0.003, b=0.003)
def forward(self, inputs, actions):
x = torch.cat((inputs, actions), 1)
x = self.linear1(x)
x = self.ln1(x)
x = F.tanh(x)
x = self.linear2(x)
x = self.ln2(x)
x = F.tanh(x)
V = self.V(x)
return V
def get_inputs():
return [torch.rand([4, 4]), torch.rand([4, 4])]
def get_init_inputs():
return [[], {'hidden_size': 4, 'num_inputs': 4, 'action_space': torch.
rand([4, 4])}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime.triton_helpers import libdevice
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_cat_0(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK: tl.
constexpr):
xnumel = 32
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 8
x1 = xindex // 8
x2 = xindex
tmp0 = x0
tl.full([1], 0, tl.int64)
tmp3 = tl.full([1], 4, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = tl.load(in_ptr0 + (4 * x1 + x0), tmp4 & xmask, eviction_policy=
'evict_last', other=0.0)
tmp6 = tmp0 >= tmp3
tl.full([1], 8, tl.int64)
tmp9 = tl.load(in_ptr1 + (4 * x1 + (-4 + x0)), tmp6 & xmask,
eviction_policy='evict_last', other=0.0)
tmp10 = tl.where(tmp4, tmp5, tmp9)
tl.store(out_ptr0 + x2, tmp10, xmask)
@triton.jit
def triton_poi_fused_native_layer_norm_1(in_ptr0, out_ptr0, out_ptr1,
xnumel, XBLOCK: tl.constexpr):
xnumel = 4
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + 4 * x0, xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr0 + (1 + 4 * x0), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr0 + (2 + 4 * x0), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (3 + 4 * x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp4 = tmp2 + tmp3
tmp6 = tmp4 + tmp5
tmp7 = 4.0
tmp8 = tmp6 / tmp7
tmp9 = tmp0 - tmp8
tmp10 = tmp9 * tmp9
tmp11 = tmp1 - tmp8
tmp12 = tmp11 * tmp11
tmp13 = tmp10 + tmp12
tmp14 = tmp3 - tmp8
tmp15 = tmp14 * tmp14
tmp16 = tmp13 + tmp15
tmp17 = tmp5 - tmp8
tmp18 = tmp17 * tmp17
tmp19 = tmp16 + tmp18
tmp20 = tmp19 / tmp7
tmp21 = 1e-05
tmp22 = tmp20 + tmp21
tmp23 = libdevice.rsqrt(tmp22)
tl.store(out_ptr0 + x0, tmp8, xmask)
tl.store(out_ptr1 + x0, tmp23, xmask)
@triton.jit
def triton_poi_fused_native_layer_norm_tanh_2(in_ptr0, in_ptr1, in_ptr2,
in_ptr3, in_ptr4, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
x0 = xindex % 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr1 + x1, xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr2 + x1, xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr3 + x0, xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr4 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 - tmp1
tmp4 = tmp2 * tmp3
tmp6 = tmp4 * tmp5
tmp8 = tmp6 + tmp7
tmp9 = libdevice.tanh(tmp8)
tl.store(out_ptr0 + x2, tmp9, xmask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7, primals_8, primals_9, primals_10, primals_11, primals_12
) = args
args.clear()
assert_size_stride(primals_1, (4, 4), (4, 1))
assert_size_stride(primals_2, (4, 4), (4, 1))
assert_size_stride(primals_3, (4, 8), (8, 1))
assert_size_stride(primals_4, (4,), (1,))
assert_size_stride(primals_5, (4,), (1,))
assert_size_stride(primals_6, (4,), (1,))
assert_size_stride(primals_7, (4, 4), (4, 1))
assert_size_stride(primals_8, (4,), (1,))
assert_size_stride(primals_9, (4,), (1,))
assert_size_stride(primals_10, (4,), (1,))
assert_size_stride(primals_11, (1, 4), (4, 1))
assert_size_stride(primals_12, (1,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 8), (8, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_cat_0[grid(32)](primals_1, primals_2, buf0, 32,
XBLOCK=32, num_warps=1, num_stages=1)
del primals_1
del primals_2
buf1 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.addmm(primals_4, buf0, reinterpret_tensor(primals_3,
(8, 4), (1, 8), 0), alpha=1, beta=1, out=buf1)
del primals_3
del primals_4
buf2 = empty_strided_cuda((4, 1), (1, 4), torch.float32)
buf3 = empty_strided_cuda((4, 1), (1, 4), torch.float32)
triton_poi_fused_native_layer_norm_1[grid(4)](buf1, buf2, buf3, 4,
XBLOCK=4, num_warps=1, num_stages=1)
buf4 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
triton_poi_fused_native_layer_norm_tanh_2[grid(16)](buf1, buf2,
buf3, primals_5, primals_6, buf4, 16, XBLOCK=16, num_warps=1,
num_stages=1)
del primals_6
buf5 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.addmm(primals_8, buf4, reinterpret_tensor(primals_7,
(4, 4), (1, 4), 0), alpha=1, beta=1, out=buf5)
del primals_8
buf6 = buf3
del buf3
buf7 = buf2
del buf2
triton_poi_fused_native_layer_norm_1[grid(4)](buf5, buf6, buf7, 4,
XBLOCK=4, num_warps=1, num_stages=1)
buf8 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
triton_poi_fused_native_layer_norm_tanh_2[grid(16)](buf5, buf6,
buf7, primals_9, primals_10, buf8, 16, XBLOCK=16, num_warps=1,
num_stages=1)
del buf6
del primals_10
buf10 = reinterpret_tensor(buf7, (4, 1), (1, 1), 0)
del buf7
extern_kernels.addmm(primals_12, buf8, reinterpret_tensor(
primals_11, (4, 1), (1, 4), 0), alpha=1, beta=1, out=buf10)
del primals_12
return (buf10, primals_5, primals_9, buf0, buf1, buf4, buf5, buf8,
primals_11, primals_7)
class CriticNew(nn.Module):
def __init__(self, hidden_size, num_inputs, action_space):
super(CriticNew, self).__init__()
self.action_space = action_space
num_outputs = action_space.shape[0]
self.linear1 = nn.Linear(num_inputs + num_outputs, hidden_size)
nn.init.normal_(self.linear1.weight, 0.0, 0.02)
self.ln1 = nn.LayerNorm(hidden_size)
self.linear2 = nn.Linear(hidden_size, hidden_size)
nn.init.normal_(self.linear2.weight, 0.0, 0.02)
self.ln2 = nn.LayerNorm(hidden_size)
self.V = nn.Linear(hidden_size, 1)
torch.nn.init.uniform_(self.V.weight, a=-0.003, b=0.003)
def forward(self, input_0, input_1):
primals_3 = self.linear1.weight
primals_4 = self.linear1.bias
primals_5 = self.ln1.weight
primals_6 = self.ln1.bias
primals_1 = self.linear2.weight
primals_8 = self.linear2.bias
primals_9 = self.ln2.weight
primals_10 = self.ln2.bias
primals_11 = self.V.weight
primals_12 = self.V.bias
primals_2 = input_0
primals_7 = input_1
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7, primals_8, primals_9,
primals_10, primals_11, primals_12])
return output[0]
| icml2019-anonymous-author/Action-Robust-Reinforcement-Learning | Critic | false | 6,852 | [
"MIT"
] | 1 | 03f0a1dd5f4a0fc5230c0ad0b41f63161bae862b | https://github.com/icml2019-anonymous-author/Action-Robust-Reinforcement-Learning/tree/03f0a1dd5f4a0fc5230c0ad0b41f63161bae862b | import torch
import torch.nn.functional as F
import torch.nn as nn
class Model(nn.Module):
def __init__(self, hidden_size, num_inputs, action_space):
super().__init__()
self.action_space = action_space
num_outputs = action_space.shape[0]
self.linear1 = nn.Linear(num_inputs + num_outputs, hidden_size)
nn.init.normal_(self.linear1.weight, 0.0, 0.02)
self.ln1 = nn.LayerNorm(hidden_size)
self.linear2 = nn.Linear(hidden_size, hidden_size)
nn.init.normal_(self.linear2.weight, 0.0, 0.02)
self.ln2 = nn.LayerNorm(hidden_size)
self.V = nn.Linear(hidden_size, 1)
torch.nn.init.uniform_(self.V.weight, a=-0.003, b=0.003)
def forward(self, inputs, actions):
x = torch.cat((inputs, actions), 1)
x = self.linear1(x)
x = self.ln1(x)
x = F.tanh(x)
x = self.linear2(x)
x = self.ln2(x)
x = F.tanh(x)
V = self.V(x)
return V
def get_inputs():
return [torch.rand([4, 4]), torch.rand([4, 4])]
def get_init_inputs():
return [[], {'hidden_size': 4, 'num_inputs': 4, 'action_space': torch.
rand([4, 4])}]
|
ScaledDotProductAttention | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/gp/cgpqjykrw2kdl4mamzzzvxwq63hpniu7qsktuwwcpeqnhbib3yz4.py
# Topologically Sorted Source Nodes: [wrapped_sqrt, scores_2], Original ATen: [aten.sqrt, aten._softmax]
# Source node to ATen node mapping:
# scores_2 => exp
# wrapped_sqrt => full_default
# Graph fragment:
# %full_default : [num_users=2] = call_function[target=torch.ops.aten.full.default](args = ([], 2.0), kwargs = {dtype: torch.float64, layout: torch.strided, device: cpu, pin_memory: False})
# %ge_scalar : [num_users=1] = call_function[target=torch.ops.aten.ge.Scalar](args = (%full_default, 0), kwargs = {})
# %scalar_tensor_default : [num_users=2] = call_function[target=torch.ops.aten.scalar_tensor.default](args = (1,), kwargs = {dtype: torch.float32, device: cuda:0, pin_memory: False})
# %neg_default : [num_users=1] = call_function[target=torch.ops.aten.neg.default](args = (%scalar_tensor_default,), kwargs = {})
# %where_self : [num_users=2] = call_function[target=torch.ops.aten.where.self](args = (%ge_scalar, %scalar_tensor_default, %neg_default), kwargs = {})
# %mul_tensor : [num_users=2] = call_function[target=torch.ops.aten.mul.Tensor](args = (%view_2, %where_self), kwargs = {})
# %amax_default : [num_users=1] = call_function[target=torch.ops.aten.amax.default](args = (%mul_tensor, [-1], True), kwargs = {})
# %sub_tensor : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%mul_tensor, %amax_default), kwargs = {})
# %mul_tensor_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%where_self, %full_default), kwargs = {})
# %div_tensor : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%sub_tensor, %mul_tensor_1), kwargs = {})
# %exp : [num_users=2] = call_function[target=torch.ops.aten.exp.default](args = (%div_tensor,), kwargs = {})
triton_poi_fused__softmax_sqrt_0 = async_compile.triton('triton_poi_fused__softmax_sqrt_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_sqrt_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_sqrt_0(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp8 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp10 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp13 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp16 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp1 = tl.full([1], 2.0, tl.float64)
tmp2 = tl.full([1], 0.0, tl.float64)
tmp3 = tmp1 >= tmp2
tmp4 = 1.0
tmp5 = -1.0
tmp6 = tl.where(tmp3, tmp4, tmp5)
tmp7 = tmp0 * tmp6
tmp9 = tmp8 * tmp6
tmp11 = tmp10 * tmp6
tmp12 = triton_helpers.maximum(tmp9, tmp11)
tmp14 = tmp13 * tmp6
tmp15 = triton_helpers.maximum(tmp12, tmp14)
tmp17 = tmp16 * tmp6
tmp18 = triton_helpers.maximum(tmp15, tmp17)
tmp19 = tmp7 - tmp18
tmp20 = tmp6.to(tl.float64)
tmp21 = tmp20 * tmp1
tmp22 = tmp21.to(tl.float32)
tmp23 = tmp19 / tmp22
tmp24 = tl_math.exp(tmp23)
tl.store(out_ptr0 + (x2), tmp24, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/fj/cfjl47pvhwbpfbvh6rfehwy5ijxc5p3zgkld2lwf3mw5bl6pbkak.py
# Topologically Sorted Source Nodes: [scores_2], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# scores_2 => div_1, sum_1
# Graph fragment:
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%exp, [-1], True), kwargs = {})
# %div_1 : [num_users=2] = call_function[target=torch.ops.aten.div.Tensor](args = (%exp, %sum_1), kwargs = {})
triton_poi_fused__softmax_1 = async_compile.triton('triton_poi_fused__softmax_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_1(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + (x2), tmp8, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, arg1_1, arg2_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg1_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg2_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((16, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [scores], Original ATen: [aten.bmm]
extern_kernels.bmm(reinterpret_tensor(arg1_1, (16, 4, 4), (16, 4, 1), 0), reinterpret_tensor(arg0_1, (16, 4, 4), (16, 1, 4), 0), out=buf0)
del arg0_1
del arg1_1
buf1 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [wrapped_sqrt, scores_2], Original ATen: [aten.sqrt, aten._softmax]
stream0 = get_raw_stream(0)
triton_poi_fused__softmax_sqrt_0.run(buf0, buf1, 256, grid=grid(256), stream=stream0)
buf2 = reinterpret_tensor(buf0, (4, 4, 4, 4), (64, 16, 4, 1), 0); del buf0 # reuse
# Topologically Sorted Source Nodes: [scores_2], Original ATen: [aten._softmax]
triton_poi_fused__softmax_1.run(buf1, buf2, 256, grid=grid(256), stream=stream0)
buf3 = reinterpret_tensor(buf1, (16, 4, 4), (16, 4, 1), 0); del buf1 # reuse
# Topologically Sorted Source Nodes: [matmul_1], Original ATen: [aten.bmm]
extern_kernels.bmm(reinterpret_tensor(buf2, (16, 4, 4), (16, 4, 1), 0), reinterpret_tensor(arg2_1, (16, 4, 4), (16, 4, 1), 0), out=buf3)
del arg2_1
return (reinterpret_tensor(buf3, (4, 4, 4, 4), (64, 16, 4, 1), 0), buf2, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
arg1_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
arg2_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([arg0_1, arg1_1, arg2_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import numpy as np
import torch.utils.data
class ScaledDotProductAttention(torch.nn.Module):
"""
Scaled, softmax attention module for Transformer as defined by
Attention(Q, K, V) on pg 4. Returns the final attention vectors as well as
the attention matrices (pairwise scores). """
def __init__(self):
super(ScaledDotProductAttention, self).__init__()
self.softmax = torch.nn.Softmax(dim=-1)
def forward(self, Q, K, V, mask=None, dropout=None):
scores = torch.matmul(Q, K.transpose(-2, -1))
scores = scores / np.sqrt(K.shape[-1])
if mask is not None:
scores = scores.masked_fill(mask == 0, -np.inf)
scores = self.softmax(scores)
if dropout is not None:
scores = dropout(scores)
return torch.matmul(scores, V), scores
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4]), torch.rand(
[4, 4, 4, 4])]
def get_init_inputs():
return [[], {}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import math as tl_math
import torch.utils.data
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused__softmax_sqrt_0(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.
constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp8 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp10 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last'
)
tmp13 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last'
)
tmp16 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last'
)
tmp1 = tl.full([1], 2.0, tl.float64)
tmp2 = tl.full([1], 0.0, tl.float64)
tmp3 = tmp1 >= tmp2
tmp4 = 1.0
tmp5 = -1.0
tmp6 = tl.where(tmp3, tmp4, tmp5)
tmp7 = tmp0 * tmp6
tmp9 = tmp8 * tmp6
tmp11 = tmp10 * tmp6
tmp12 = triton_helpers.maximum(tmp9, tmp11)
tmp14 = tmp13 * tmp6
tmp15 = triton_helpers.maximum(tmp12, tmp14)
tmp17 = tmp16 * tmp6
tmp18 = triton_helpers.maximum(tmp15, tmp17)
tmp19 = tmp7 - tmp18
tmp20 = tmp6.to(tl.float64)
tmp21 = tmp20 * tmp1
tmp22 = tmp21.to(tl.float32)
tmp23 = tmp19 / tmp22
tmp24 = tl_math.exp(tmp23)
tl.store(out_ptr0 + x2, tmp24, xmask)
@triton.jit
def triton_poi_fused__softmax_1(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + x2, tmp8, xmask)
def call(args):
arg0_1, arg1_1, arg2_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg1_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg2_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((16, 4, 4), (16, 4, 1), torch.float32)
extern_kernels.bmm(reinterpret_tensor(arg1_1, (16, 4, 4), (16, 4, 1
), 0), reinterpret_tensor(arg0_1, (16, 4, 4), (16, 1, 4), 0),
out=buf0)
del arg0_1
del arg1_1
buf1 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused__softmax_sqrt_0[grid(256)](buf0, buf1, 256, XBLOCK
=256, num_warps=4, num_stages=1)
buf2 = reinterpret_tensor(buf0, (4, 4, 4, 4), (64, 16, 4, 1), 0)
del buf0
triton_poi_fused__softmax_1[grid(256)](buf1, buf2, 256, XBLOCK=256,
num_warps=4, num_stages=1)
buf3 = reinterpret_tensor(buf1, (16, 4, 4), (16, 4, 1), 0)
del buf1
extern_kernels.bmm(reinterpret_tensor(buf2, (16, 4, 4), (16, 4, 1),
0), reinterpret_tensor(arg2_1, (16, 4, 4), (16, 4, 1), 0), out=buf3
)
del arg2_1
return reinterpret_tensor(buf3, (4, 4, 4, 4), (64, 16, 4, 1), 0), buf2
class ScaledDotProductAttentionNew(torch.nn.Module):
"""
Scaled, softmax attention module for Transformer as defined by
Attention(Q, K, V) on pg 4. Returns the final attention vectors as well as
the attention matrices (pairwise scores). """
def __init__(self):
super(ScaledDotProductAttentionNew, self).__init__()
self.softmax = torch.nn.Softmax(dim=-1)
def forward(self, input_0, input_1, input_2):
arg0_1 = input_0
arg1_1 = input_1
arg2_1 = input_2
output = call([arg0_1, arg1_1, arg2_1])
return output[0], output[1]
| icemansina/protein-transformer | ScaledDotProductAttention | false | 6,853 | [
"BSD-3-Clause"
] | 1 | 4e73b17f2a4b89ba1a9f6703976d1a31b7a8a5eb | https://github.com/icemansina/protein-transformer/tree/4e73b17f2a4b89ba1a9f6703976d1a31b7a8a5eb | import torch
import numpy as np
import torch.utils.data
class Model(torch.nn.Module):
"""
Scaled, softmax attention module for Transformer as defined by
Attention(Q, K, V) on pg 4. Returns the final attention vectors as well as
the attention matrices (pairwise scores). """
def __init__(self):
super().__init__()
self.softmax = torch.nn.Softmax(dim=-1)
def forward(self, Q, K, V, mask=None, dropout=None):
scores = torch.matmul(Q, K.transpose(-2, -1))
scores = scores / np.sqrt(K.shape[-1])
if mask is not None:
scores = scores.masked_fill(mask == 0, -np.inf)
scores = self.softmax(scores)
if dropout is not None:
scores = dropout(scores)
return torch.matmul(scores, V), scores
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4]), torch.rand(
[4, 4, 4, 4])]
def get_init_inputs():
return []
|
Actor | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/v7/cv7zazascu4rpkkwoxbiwk6c2le2e6wshdhae73bmaoapelvwguv.py
# Topologically Sorted Source Nodes: [a], Original ATen: [aten.relu, aten.threshold_backward]
# Source node to ATen node mapping:
# a => relu
# Graph fragment:
# %relu : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%view_1,), kwargs = {})
# %le_1 : [num_users=1] = call_function[target=torch.ops.aten.le.Scalar](args = (%relu, 0), kwargs = {})
triton_poi_fused_relu_threshold_backward_0 = async_compile.triton('triton_poi_fused_relu_threshold_backward_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[1024],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*i1', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_relu_threshold_backward_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_relu_threshold_backward_0(in_out_ptr0, in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 1024
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 16
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(in_out_ptr0 + (x2), tmp4, xmask)
tl.store(out_ptr0 + (x2), tmp6, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/cp/ccp5m5apf7ka2skqyfxhf2df54c52qocprpycry7jrzoptyjvbti.py
# Topologically Sorted Source Nodes: [tanh, a_2], Original ATen: [aten.tanh, aten.mul]
# Source node to ATen node mapping:
# a_2 => mul
# tanh => tanh
# Graph fragment:
# %tanh : [num_users=1] = call_function[target=torch.ops.aten.tanh.default](args = (%view_5,), kwargs = {})
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%tanh, 4), kwargs = {})
triton_poi_fused_mul_tanh_1 = async_compile.triton('triton_poi_fused_mul_tanh_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_mul_tanh_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_mul_tanh_1(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (x0), xmask)
tmp1 = libdevice.tanh(tmp0)
tmp2 = 4.0
tmp3 = tmp1 * tmp2
tl.store(out_ptr0 + (x0), tmp3, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7 = args
args.clear()
assert_size_stride(primals_1, (16, 4), (4, 1))
assert_size_stride(primals_2, (16, ), (1, ))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_4, (16, 16), (16, 1))
assert_size_stride(primals_5, (16, ), (1, ))
assert_size_stride(primals_6, (4, 16), (16, 1))
assert_size_stride(primals_7, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 16), (16, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(primals_3, (64, 4), (4, 1), 0), reinterpret_tensor(primals_1, (4, 16), (1, 4), 0), out=buf0)
del primals_1
buf1 = reinterpret_tensor(buf0, (4, 4, 4, 16), (256, 64, 16, 1), 0); del buf0 # reuse
buf7 = empty_strided_cuda((4, 4, 4, 16), (256, 64, 16, 1), torch.bool)
# Topologically Sorted Source Nodes: [a], Original ATen: [aten.relu, aten.threshold_backward]
stream0 = get_raw_stream(0)
triton_poi_fused_relu_threshold_backward_0.run(buf1, primals_2, buf7, 1024, grid=grid(1024), stream=stream0)
del primals_2
buf2 = empty_strided_cuda((64, 16), (16, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(buf1, (64, 16), (16, 1), 0), reinterpret_tensor(primals_4, (16, 16), (1, 16), 0), out=buf2)
buf3 = reinterpret_tensor(buf2, (4, 4, 4, 16), (256, 64, 16, 1), 0); del buf2 # reuse
buf6 = empty_strided_cuda((4, 4, 4, 16), (256, 64, 16, 1), torch.bool)
# Topologically Sorted Source Nodes: [a_1], Original ATen: [aten.relu, aten.threshold_backward]
triton_poi_fused_relu_threshold_backward_0.run(buf3, primals_5, buf6, 1024, grid=grid(1024), stream=stream0)
del primals_5
buf4 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [linear_2], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_7, reinterpret_tensor(buf3, (64, 16), (16, 1), 0), reinterpret_tensor(primals_6, (16, 4), (1, 16), 0), alpha=1, beta=1, out=buf4)
del primals_7
buf5 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [tanh, a_2], Original ATen: [aten.tanh, aten.mul]
triton_poi_fused_mul_tanh_1.run(buf4, buf5, 256, grid=grid(256), stream=stream0)
return (buf5, reinterpret_tensor(primals_3, (64, 4), (4, 1), 0), reinterpret_tensor(buf1, (64, 16), (16, 1), 0), reinterpret_tensor(buf3, (64, 16), (16, 1), 0), buf4, primals_6, buf6, primals_4, buf7, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((16, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((16, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((16, 16), (16, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((16, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((4, 16), (16, 1), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch as t
import torch.nn as nn
class Actor(nn.Module):
def __init__(self, state_dim, action_dim, action_range):
super().__init__()
self.fc1 = nn.Linear(state_dim, 16)
self.fc2 = nn.Linear(16, 16)
self.fc3 = nn.Linear(16, action_dim)
self.action_range = action_range
def forward(self, state):
a = t.relu(self.fc1(state))
a = t.relu(self.fc2(a))
a = t.tanh(self.fc3(a)) * self.action_range
return a
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'state_dim': 4, 'action_dim': 4, 'action_range': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import libdevice
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_relu_threshold_backward_0(in_out_ptr0, in_ptr0,
out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 1024
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 16
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(in_out_ptr0 + x2, tmp4, xmask)
tl.store(out_ptr0 + x2, tmp6, xmask)
@triton.jit
def triton_poi_fused_mul_tanh_1(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + x0, xmask)
tmp1 = libdevice.tanh(tmp0)
tmp2 = 4.0
tmp3 = tmp1 * tmp2
tl.store(out_ptr0 + x0, tmp3, xmask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7) = args
args.clear()
assert_size_stride(primals_1, (16, 4), (4, 1))
assert_size_stride(primals_2, (16,), (1,))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_4, (16, 16), (16, 1))
assert_size_stride(primals_5, (16,), (1,))
assert_size_stride(primals_6, (4, 16), (16, 1))
assert_size_stride(primals_7, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 16), (16, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_3, (64, 4), (4, 1), 0),
reinterpret_tensor(primals_1, (4, 16), (1, 4), 0), out=buf0)
del primals_1
buf1 = reinterpret_tensor(buf0, (4, 4, 4, 16), (256, 64, 16, 1), 0)
del buf0
buf7 = empty_strided_cuda((4, 4, 4, 16), (256, 64, 16, 1), torch.bool)
get_raw_stream(0)
triton_poi_fused_relu_threshold_backward_0[grid(1024)](buf1,
primals_2, buf7, 1024, XBLOCK=128, num_warps=4, num_stages=1)
del primals_2
buf2 = empty_strided_cuda((64, 16), (16, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(buf1, (64, 16), (16, 1), 0),
reinterpret_tensor(primals_4, (16, 16), (1, 16), 0), out=buf2)
buf3 = reinterpret_tensor(buf2, (4, 4, 4, 16), (256, 64, 16, 1), 0)
del buf2
buf6 = empty_strided_cuda((4, 4, 4, 16), (256, 64, 16, 1), torch.bool)
triton_poi_fused_relu_threshold_backward_0[grid(1024)](buf3,
primals_5, buf6, 1024, XBLOCK=128, num_warps=4, num_stages=1)
del primals_5
buf4 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
extern_kernels.addmm(primals_7, reinterpret_tensor(buf3, (64, 16),
(16, 1), 0), reinterpret_tensor(primals_6, (16, 4), (1, 16), 0),
alpha=1, beta=1, out=buf4)
del primals_7
buf5 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
triton_poi_fused_mul_tanh_1[grid(256)](buf4, buf5, 256, XBLOCK=256,
num_warps=4, num_stages=1)
return buf5, reinterpret_tensor(primals_3, (64, 4), (4, 1), 0
), reinterpret_tensor(buf1, (64, 16), (16, 1), 0), reinterpret_tensor(
buf3, (64, 16), (16, 1), 0), buf4, primals_6, buf6, primals_4, buf7
class ActorNew(nn.Module):
def __init__(self, state_dim, action_dim, action_range):
super().__init__()
self.fc1 = nn.Linear(state_dim, 16)
self.fc2 = nn.Linear(16, 16)
self.fc3 = nn.Linear(16, action_dim)
self.action_range = action_range
def forward(self, input_0):
primals_1 = self.fc1.weight
primals_2 = self.fc1.bias
primals_4 = self.fc2.weight
primals_5 = self.fc2.bias
primals_6 = self.fc3.weight
primals_7 = self.fc3.bias
primals_3 = input_0
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7])
return output[0]
| ikamensh/machin | Actor | false | 6,854 | [
"MIT"
] | 1 | af7b423c47bc1412530cf6c96c11bd3af9b3e239 | https://github.com/ikamensh/machin/tree/af7b423c47bc1412530cf6c96c11bd3af9b3e239 | import torch
import torch as t
import torch.nn as nn
class Model(nn.Module):
def __init__(self, state_dim, action_dim, action_range):
super().__init__()
self.fc1 = nn.Linear(state_dim, 16)
self.fc2 = nn.Linear(16, 16)
self.fc3 = nn.Linear(16, action_dim)
self.action_range = action_range
def forward(self, state):
a = t.relu(self.fc1(state))
a = t.relu(self.fc2(a))
a = t.tanh(self.fc3(a)) * self.action_range
return a
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4, 4, 4]
|
MultiHeadedAttention | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/dk/cdk4odz276xorciau5ehgl7f3s2mgkf3hrye6xep6kzubczdeqqy.py
# Topologically Sorted Source Nodes: [matmul], Original ATen: [aten.clone]
# Source node to ATen node mapping:
# matmul => clone
# Graph fragment:
# %clone : [num_users=1] = call_function[target=torch.ops.aten.clone.default](args = (%expand,), kwargs = {memory_format: torch.contiguous_format})
triton_poi_fused_clone_0 = async_compile.triton('triton_poi_fused_clone_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16, 4], tile_hint=TileHint.DEFAULT,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clone_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clone_0(in_ptr0, in_ptr1, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = (yindex // 4)
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + (4*x2) + (16*y1)), xmask & ymask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (y0), ymask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(out_ptr0 + (x2 + (4*y3)), tmp2, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/bs/cbsluabtq7ll426nybkislhh3cajm6f7ggrxam362hohynwnvtk6.py
# Topologically Sorted Source Nodes: [eq], Original ATen: [aten.eq]
# Source node to ATen node mapping:
# eq => eq
# Graph fragment:
# %eq : [num_users=2] = call_function[target=torch.ops.aten.eq.Scalar](args = (%unsqueeze, 0), kwargs = {})
triton_poi_fused_eq_1 = async_compile.triton('triton_poi_fused_eq_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*i1', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_eq_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_eq_1(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (x0), xmask)
tmp1 = 0.0
tmp2 = tmp0 == tmp1
tl.store(out_ptr0 + (x0), tmp2, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/6v/c6varcicuq4byhpi2ez4zfksc6c5naym336xenou4witslx53n6e.py
# Topologically Sorted Source Nodes: [scores, scores_1, p_attn], Original ATen: [aten.div, aten.masked_fill, aten._softmax]
# Source node to ATen node mapping:
# p_attn => amax, exp, sub, sum_1
# scores => div
# scores_1 => full_default, where
# Graph fragment:
# %div : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%view_11, 1.0), kwargs = {})
# %full_default : [num_users=1] = call_function[target=torch.ops.aten.full.default](args = ([], -1000000000.0), kwargs = {dtype: torch.float32, layout: torch.strided, device: cuda:0, pin_memory: False})
# %where : [num_users=2] = call_function[target=torch.ops.aten.where.self](args = (%eq, %full_default, %div), kwargs = {})
# %amax : [num_users=1] = call_function[target=torch.ops.aten.amax.default](args = (%where, [-1], True), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%where, %amax), kwargs = {})
# %exp : [num_users=2] = call_function[target=torch.ops.aten.exp.default](args = (%sub,), kwargs = {})
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%exp, [-1], True), kwargs = {})
triton_poi_fused__softmax_div_masked_fill_2 = async_compile.triton('triton_poi_fused__softmax_div_masked_fill_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*i1', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_div_masked_fill_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 8, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_div_masked_fill_2(in_ptr0, in_ptr1, out_ptr0, out_ptr1, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 4
x2 = (xindex // 16)
x3 = xindex
tmp0 = tl.load(in_ptr0 + ((4*x0) + (16*x2)), xmask, eviction_policy='evict_last').to(tl.int1)
tmp1 = tl.load(in_ptr1 + (4*x3), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (1 + (4*x0) + (16*x2)), xmask, eviction_policy='evict_last').to(tl.int1)
tmp7 = tl.load(in_ptr1 + (1 + (4*x3)), xmask, eviction_policy='evict_last')
tmp11 = tl.load(in_ptr0 + (2 + (4*x0) + (16*x2)), xmask, eviction_policy='evict_last').to(tl.int1)
tmp12 = tl.load(in_ptr1 + (2 + (4*x3)), xmask, eviction_policy='evict_last')
tmp16 = tl.load(in_ptr0 + (3 + (4*x0) + (16*x2)), xmask, eviction_policy='evict_last').to(tl.int1)
tmp17 = tl.load(in_ptr1 + (3 + (4*x3)), xmask, eviction_policy='evict_last')
tmp2 = 1.0
tmp3 = tmp1 * tmp2
tmp4 = -1000000000.0
tmp5 = tl.where(tmp0, tmp4, tmp3)
tmp8 = tmp7 * tmp2
tmp9 = tl.where(tmp6, tmp4, tmp8)
tmp10 = triton_helpers.maximum(tmp5, tmp9)
tmp13 = tmp12 * tmp2
tmp14 = tl.where(tmp11, tmp4, tmp13)
tmp15 = triton_helpers.maximum(tmp10, tmp14)
tmp18 = tmp17 * tmp2
tmp19 = tl.where(tmp16, tmp4, tmp18)
tmp20 = triton_helpers.maximum(tmp15, tmp19)
tmp21 = tmp5 - tmp20
tmp22 = tl_math.exp(tmp21)
tmp23 = tmp9 - tmp20
tmp24 = tl_math.exp(tmp23)
tmp25 = tmp22 + tmp24
tmp26 = tmp14 - tmp20
tmp27 = tl_math.exp(tmp26)
tmp28 = tmp25 + tmp27
tmp29 = tmp19 - tmp20
tmp30 = tl_math.exp(tmp29)
tmp31 = tmp28 + tmp30
tl.store(out_ptr0 + (x3), tmp20, xmask)
tl.store(out_ptr1 + (x3), tmp31, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/fb/cfb6x4ausnhkt7gycxsexlmn75uyb6haot5nthwqf5hummcazrv4.py
# Topologically Sorted Source Nodes: [scores, scores_1, p_attn], Original ATen: [aten.div, aten.masked_fill, aten._softmax]
# Source node to ATen node mapping:
# p_attn => amax, div_1, exp, sub
# scores => div
# scores_1 => full_default, where
# Graph fragment:
# %div : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%view_11, 1.0), kwargs = {})
# %full_default : [num_users=1] = call_function[target=torch.ops.aten.full.default](args = ([], -1000000000.0), kwargs = {dtype: torch.float32, layout: torch.strided, device: cuda:0, pin_memory: False})
# %where : [num_users=2] = call_function[target=torch.ops.aten.where.self](args = (%eq, %full_default, %div), kwargs = {})
# %amax : [num_users=1] = call_function[target=torch.ops.aten.amax.default](args = (%where, [-1], True), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%where, %amax), kwargs = {})
# %exp : [num_users=2] = call_function[target=torch.ops.aten.exp.default](args = (%sub,), kwargs = {})
# %div_1 : [num_users=2] = call_function[target=torch.ops.aten.div.Tensor](args = (%exp, %sum_1), kwargs = {})
triton_poi_fused__softmax_div_masked_fill_3 = async_compile.triton('triton_poi_fused__softmax_div_masked_fill_3', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*i1', 2: '*fp32', 3: '*fp32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_div_masked_fill_3', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 4, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_div_masked_fill_3(in_out_ptr0, in_ptr0, in_ptr1, in_ptr2, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = (xindex // 64)
x4 = xindex % 16
x5 = xindex
x6 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x4 + (16*x3)), xmask, eviction_policy='evict_last').to(tl.int1)
tmp1 = tl.load(in_out_ptr0 + (x5), xmask)
tmp6 = tl.load(in_ptr1 + (x6), xmask, eviction_policy='evict_last')
tmp9 = tl.load(in_ptr2 + (x6), xmask, eviction_policy='evict_last')
tmp2 = 1.0
tmp3 = tmp1 * tmp2
tmp4 = -1000000000.0
tmp5 = tl.where(tmp0, tmp4, tmp3)
tmp7 = tmp5 - tmp6
tmp8 = tl_math.exp(tmp7)
tmp10 = tmp8 / tmp9
tl.store(in_out_ptr0 + (x5), tmp10, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/6t/c6t5a5ere3lqjiu7zh3uu4oxmpdoujdaqqmeunxqapgzo4m74uav.py
# Topologically Sorted Source Nodes: [contiguous], Original ATen: [aten.clone]
# Source node to ATen node mapping:
# contiguous => clone_4
# Graph fragment:
# %clone_4 : [num_users=1] = call_function[target=torch.ops.aten.clone.default](args = (%permute_7,), kwargs = {memory_format: torch.contiguous_format})
triton_poi_fused_clone_4 = async_compile.triton('triton_poi_fused_clone_4', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16, 4], tile_hint=TileHint.SQUARE,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clone_4', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clone_4(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = (yindex // 4)
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + (4*x2) + (16*y1)), xmask & ymask, eviction_policy='evict_last')
tl.store(out_ptr0 + (x2 + (4*y3)), tmp0, xmask & ymask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9, primals_10, primals_11, primals_12 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_2, (4, 4), (4, 1))
assert_size_stride(primals_3, (4, ), (1, ))
assert_size_stride(primals_4, (4, 4), (4, 1))
assert_size_stride(primals_5, (4, ), (1, ))
assert_size_stride(primals_6, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_7, (4, 4), (4, 1))
assert_size_stride(primals_8, (4, ), (1, ))
assert_size_stride(primals_9, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_10, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_11, (4, 4), (4, 1))
assert_size_stride(primals_12, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(primals_1, (16, 4), (4, 1), 0), reinterpret_tensor(primals_2, (4, 4), (1, 4), 0), out=buf0)
del primals_2
buf1 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(primals_6, (16, 4), (4, 1), 0), reinterpret_tensor(primals_4, (4, 4), (1, 4), 0), out=buf1)
del primals_4
buf2 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(primals_9, (16, 4), (4, 1), 0), reinterpret_tensor(primals_7, (4, 4), (1, 4), 0), out=buf2)
del primals_7
buf3 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 1), torch.float32)
# Topologically Sorted Source Nodes: [matmul], Original ATen: [aten.clone]
stream0 = get_raw_stream(0)
triton_poi_fused_clone_0.run(buf0, primals_3, buf3, 16, 4, grid=grid(16, 4), stream=stream0)
del primals_3
buf4 = reinterpret_tensor(buf0, (4, 4, 1, 4), (16, 4, 4, 1), 0); del buf0 # reuse
# Topologically Sorted Source Nodes: [matmul], Original ATen: [aten.clone]
triton_poi_fused_clone_0.run(buf1, primals_5, buf4, 16, 4, grid=grid(16, 4), stream=stream0)
del primals_5
buf5 = empty_strided_cuda((16, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [matmul], Original ATen: [aten.bmm]
extern_kernels.bmm(reinterpret_tensor(buf3, (16, 4, 1), (4, 1, 0), 0), reinterpret_tensor(buf4, (16, 1, 4), (4, 0, 1), 0), out=buf5)
buf6 = empty_strided_cuda((4, 1, 4, 4), (16, 16, 4, 1), torch.bool)
# Topologically Sorted Source Nodes: [eq], Original ATen: [aten.eq]
triton_poi_fused_eq_1.run(primals_10, buf6, 64, grid=grid(64), stream=stream0)
del primals_10
buf7 = reinterpret_tensor(buf1, (4, 4, 4, 1), (16, 4, 1, 64), 0); del buf1 # reuse
buf8 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 64), torch.float32)
# Topologically Sorted Source Nodes: [scores, scores_1, p_attn], Original ATen: [aten.div, aten.masked_fill, aten._softmax]
triton_poi_fused__softmax_div_masked_fill_2.run(buf6, buf5, buf7, buf8, 64, grid=grid(64), stream=stream0)
buf9 = reinterpret_tensor(buf5, (4, 4, 4, 4), (64, 16, 4, 1), 0); del buf5 # reuse
# Topologically Sorted Source Nodes: [scores, scores_1, p_attn], Original ATen: [aten.div, aten.masked_fill, aten._softmax]
triton_poi_fused__softmax_div_masked_fill_3.run(buf9, buf6, buf7, buf8, 256, grid=grid(256), stream=stream0)
buf10 = reinterpret_tensor(buf8, (4, 4, 4, 1), (16, 4, 1, 1), 0); del buf8 # reuse
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.clone]
triton_poi_fused_clone_0.run(buf2, primals_8, buf10, 16, 4, grid=grid(16, 4), stream=stream0)
del primals_8
buf11 = reinterpret_tensor(buf2, (16, 4, 1), (4, 1, 1), 0); del buf2 # reuse
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.bmm]
extern_kernels.bmm(reinterpret_tensor(buf9, (16, 4, 4), (16, 4, 1), 0), reinterpret_tensor(buf10, (16, 4, 1), (4, 1, 0), 0), out=buf11)
buf12 = reinterpret_tensor(buf7, (4, 4, 4, 1), (16, 4, 1, 1), 0); del buf7 # reuse
# Topologically Sorted Source Nodes: [contiguous], Original ATen: [aten.clone]
triton_poi_fused_clone_4.run(buf11, buf12, 16, 4, grid=grid(16, 4), stream=stream0)
buf13 = reinterpret_tensor(buf11, (16, 4), (4, 1), 0); del buf11 # reuse
# Topologically Sorted Source Nodes: [linear_3], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_12, reinterpret_tensor(buf12, (16, 4), (4, 1), 0), reinterpret_tensor(primals_11, (4, 4), (1, 4), 0), alpha=1, beta=1, out=buf13)
del primals_12
return (reinterpret_tensor(buf13, (4, 4, 4), (16, 4, 1), 0), reinterpret_tensor(primals_1, (16, 4), (4, 1), 0), reinterpret_tensor(primals_6, (16, 4), (4, 1), 0), reinterpret_tensor(primals_9, (16, 4), (4, 1), 0), buf6, buf9, reinterpret_tensor(buf12, (16, 4), (4, 1), 0), primals_11, reinterpret_tensor(buf10, (16, 1, 4), (4, 1, 1), 0), reinterpret_tensor(buf3, (16, 1, 4), (4, 1, 1), 0), reinterpret_tensor(buf4, (16, 4, 1), (4, 1, 4), 0), )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_8 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_9 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_10 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_11 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_12 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9, primals_10, primals_11, primals_12])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import math
import torch
import torch.nn as nn
import torch.nn.functional as F
class MultiHeadedAttention(nn.Module):
def __init__(self, num_head, d_model, dropout=0.1):
super(MultiHeadedAttention, self).__init__()
assert d_model % num_head == 0
self.d_k = d_model // num_head
self.h = num_head
self.linear_key = nn.Linear(d_model, d_model)
self.linear_value = nn.Linear(d_model, d_model)
self.linear_query = nn.Linear(d_model, d_model)
self.linear_out = nn.Linear(d_model, d_model)
self.dropout = nn.Dropout(p=dropout)
def attention(self, query, key, value, mask, dropout=None):
d_k = query.size(-1)
scores = torch.matmul(query, key.transpose(-2, -1)) / math.sqrt(d_k)
scores = scores.masked_fill(mask == 0, -1000000000.0)
p_attn = F.softmax(scores, dim=-1)
if dropout is not None:
p_attn = dropout(p_attn)
return torch.matmul(p_attn, value), p_attn
def forward(self, query, key, value, mask):
nbatches = query.size(0)
query = self.linear_query(query).view(nbatches, -1, self.h, self.d_k
).transpose(1, 2)
key = self.linear_key(key).view(nbatches, -1, self.h, self.d_k
).transpose(1, 2)
value = self.linear_value(value).view(nbatches, -1, self.h, self.d_k
).transpose(1, 2)
mask = mask.unsqueeze(1)
x, _attn = self.attention(query, key, value, mask, dropout=self.dropout
)
x = x.transpose(1, 2).contiguous().view(nbatches, -1, self.h * self.d_k
)
return self.linear_out(x)
def get_inputs():
return [torch.rand([4, 4, 4]), torch.rand([4, 4, 4]), torch.rand([4, 4,
4]), torch.rand([4, 4, 4])]
def get_init_inputs():
return [[], {'num_head': 4, 'd_model': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import math as tl_math
import math
import torch.nn as nn
import torch.nn.functional as F
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_clone_0(in_ptr0, in_ptr1, out_ptr0, ynumel, xnumel,
YBLOCK: tl.constexpr, XBLOCK: tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = yindex // 4
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + 4 * x2 + 16 * y1), xmask & ymask,
eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + y0, ymask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(out_ptr0 + (x2 + 4 * y3), tmp2, xmask & ymask)
@triton.jit
def triton_poi_fused_eq_1(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + x0, xmask)
tmp1 = 0.0
tmp2 = tmp0 == tmp1
tl.store(out_ptr0 + x0, tmp2, xmask)
@triton.jit
def triton_poi_fused__softmax_div_masked_fill_2(in_ptr0, in_ptr1, out_ptr0,
out_ptr1, xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 4
x2 = xindex // 16
x3 = xindex
tmp0 = tl.load(in_ptr0 + (4 * x0 + 16 * x2), xmask, eviction_policy=
'evict_last').to(tl.int1)
tmp1 = tl.load(in_ptr1 + 4 * x3, xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (1 + 4 * x0 + 16 * x2), xmask, eviction_policy
='evict_last').to(tl.int1)
tmp7 = tl.load(in_ptr1 + (1 + 4 * x3), xmask, eviction_policy='evict_last')
tmp11 = tl.load(in_ptr0 + (2 + 4 * x0 + 16 * x2), xmask,
eviction_policy='evict_last').to(tl.int1)
tmp12 = tl.load(in_ptr1 + (2 + 4 * x3), xmask, eviction_policy='evict_last'
)
tmp16 = tl.load(in_ptr0 + (3 + 4 * x0 + 16 * x2), xmask,
eviction_policy='evict_last').to(tl.int1)
tmp17 = tl.load(in_ptr1 + (3 + 4 * x3), xmask, eviction_policy='evict_last'
)
tmp2 = 1.0
tmp3 = tmp1 * tmp2
tmp4 = -1000000000.0
tmp5 = tl.where(tmp0, tmp4, tmp3)
tmp8 = tmp7 * tmp2
tmp9 = tl.where(tmp6, tmp4, tmp8)
tmp10 = triton_helpers.maximum(tmp5, tmp9)
tmp13 = tmp12 * tmp2
tmp14 = tl.where(tmp11, tmp4, tmp13)
tmp15 = triton_helpers.maximum(tmp10, tmp14)
tmp18 = tmp17 * tmp2
tmp19 = tl.where(tmp16, tmp4, tmp18)
tmp20 = triton_helpers.maximum(tmp15, tmp19)
tmp21 = tmp5 - tmp20
tmp22 = tl_math.exp(tmp21)
tmp23 = tmp9 - tmp20
tmp24 = tl_math.exp(tmp23)
tmp25 = tmp22 + tmp24
tmp26 = tmp14 - tmp20
tmp27 = tl_math.exp(tmp26)
tmp28 = tmp25 + tmp27
tmp29 = tmp19 - tmp20
tmp30 = tl_math.exp(tmp29)
tmp31 = tmp28 + tmp30
tl.store(out_ptr0 + x3, tmp20, xmask)
tl.store(out_ptr1 + x3, tmp31, xmask)
@triton.jit
def triton_poi_fused__softmax_div_masked_fill_3(in_out_ptr0, in_ptr0,
in_ptr1, in_ptr2, xnumel, XBLOCK: tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex // 64
x4 = xindex % 16
x5 = xindex
x6 = xindex // 4
tmp0 = tl.load(in_ptr0 + (x4 + 16 * x3), xmask, eviction_policy=
'evict_last').to(tl.int1)
tmp1 = tl.load(in_out_ptr0 + x5, xmask)
tmp6 = tl.load(in_ptr1 + x6, xmask, eviction_policy='evict_last')
tmp9 = tl.load(in_ptr2 + x6, xmask, eviction_policy='evict_last')
tmp2 = 1.0
tmp3 = tmp1 * tmp2
tmp4 = -1000000000.0
tmp5 = tl.where(tmp0, tmp4, tmp3)
tmp7 = tmp5 - tmp6
tmp8 = tl_math.exp(tmp7)
tmp10 = tmp8 / tmp9
tl.store(in_out_ptr0 + x5, tmp10, xmask)
@triton.jit
def triton_poi_fused_clone_4(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK: tl.
constexpr, XBLOCK: tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = yindex // 4
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + 4 * x2 + 16 * y1), xmask & ymask,
eviction_policy='evict_last')
tl.store(out_ptr0 + (x2 + 4 * y3), tmp0, xmask & ymask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7, primals_8, primals_9, primals_10, primals_11, primals_12
) = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_2, (4, 4), (4, 1))
assert_size_stride(primals_3, (4,), (1,))
assert_size_stride(primals_4, (4, 4), (4, 1))
assert_size_stride(primals_5, (4,), (1,))
assert_size_stride(primals_6, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_7, (4, 4), (4, 1))
assert_size_stride(primals_8, (4,), (1,))
assert_size_stride(primals_9, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_10, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_11, (4, 4), (4, 1))
assert_size_stride(primals_12, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_1, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_2, (4, 4), (1, 4), 0), out=buf0)
del primals_2
buf1 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_6, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_4, (4, 4), (1, 4), 0), out=buf1)
del primals_4
buf2 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_9, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_7, (4, 4), (1, 4), 0), out=buf2)
del primals_7
buf3 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_clone_0[grid(16, 4)](buf0, primals_3, buf3, 16, 4,
XBLOCK=2, YBLOCK=16, num_warps=1, num_stages=1)
del primals_3
buf4 = reinterpret_tensor(buf0, (4, 4, 1, 4), (16, 4, 4, 1), 0)
del buf0
triton_poi_fused_clone_0[grid(16, 4)](buf1, primals_5, buf4, 16, 4,
XBLOCK=2, YBLOCK=16, num_warps=1, num_stages=1)
del primals_5
buf5 = empty_strided_cuda((16, 4, 4), (16, 4, 1), torch.float32)
extern_kernels.bmm(reinterpret_tensor(buf3, (16, 4, 1), (4, 1, 0),
0), reinterpret_tensor(buf4, (16, 1, 4), (4, 0, 1), 0), out=buf5)
buf6 = empty_strided_cuda((4, 1, 4, 4), (16, 16, 4, 1), torch.bool)
triton_poi_fused_eq_1[grid(64)](primals_10, buf6, 64, XBLOCK=64,
num_warps=1, num_stages=1)
del primals_10
buf7 = reinterpret_tensor(buf1, (4, 4, 4, 1), (16, 4, 1, 64), 0)
del buf1
buf8 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 64), torch.float32)
triton_poi_fused__softmax_div_masked_fill_2[grid(64)](buf6, buf5,
buf7, buf8, 64, XBLOCK=64, num_warps=1, num_stages=1)
buf9 = reinterpret_tensor(buf5, (4, 4, 4, 4), (64, 16, 4, 1), 0)
del buf5
triton_poi_fused__softmax_div_masked_fill_3[grid(256)](buf9, buf6,
buf7, buf8, 256, XBLOCK=256, num_warps=4, num_stages=1)
buf10 = reinterpret_tensor(buf8, (4, 4, 4, 1), (16, 4, 1, 1), 0)
del buf8
triton_poi_fused_clone_0[grid(16, 4)](buf2, primals_8, buf10, 16, 4,
XBLOCK=2, YBLOCK=16, num_warps=1, num_stages=1)
del primals_8
buf11 = reinterpret_tensor(buf2, (16, 4, 1), (4, 1, 1), 0)
del buf2
extern_kernels.bmm(reinterpret_tensor(buf9, (16, 4, 4), (16, 4, 1),
0), reinterpret_tensor(buf10, (16, 4, 1), (4, 1, 0), 0), out=buf11)
buf12 = reinterpret_tensor(buf7, (4, 4, 4, 1), (16, 4, 1, 1), 0)
del buf7
triton_poi_fused_clone_4[grid(16, 4)](buf11, buf12, 16, 4, XBLOCK=4,
YBLOCK=16, num_warps=1, num_stages=1)
buf13 = reinterpret_tensor(buf11, (16, 4), (4, 1), 0)
del buf11
extern_kernels.addmm(primals_12, reinterpret_tensor(buf12, (16, 4),
(4, 1), 0), reinterpret_tensor(primals_11, (4, 4), (1, 4), 0),
alpha=1, beta=1, out=buf13)
del primals_12
return reinterpret_tensor(buf13, (4, 4, 4), (16, 4, 1), 0
), reinterpret_tensor(primals_1, (16, 4), (4, 1), 0
), reinterpret_tensor(primals_6, (16, 4), (4, 1), 0
), reinterpret_tensor(primals_9, (16, 4), (4, 1), 0
), buf6, buf9, reinterpret_tensor(buf12, (16, 4), (4, 1), 0
), primals_11, reinterpret_tensor(buf10, (16, 1, 4), (4, 1, 1), 0
), reinterpret_tensor(buf3, (16, 1, 4), (4, 1, 1), 0
), reinterpret_tensor(buf4, (16, 4, 1), (4, 1, 4), 0)
class MultiHeadedAttentionNew(nn.Module):
def __init__(self, num_head, d_model, dropout=0.1):
super(MultiHeadedAttentionNew, self).__init__()
assert d_model % num_head == 0
self.d_k = d_model // num_head
self.h = num_head
self.linear_key = nn.Linear(d_model, d_model)
self.linear_value = nn.Linear(d_model, d_model)
self.linear_query = nn.Linear(d_model, d_model)
self.linear_out = nn.Linear(d_model, d_model)
self.dropout = nn.Dropout(p=dropout)
def attention(self, query, key, value, mask, dropout=None):
d_k = query.size(-1)
scores = torch.matmul(query, key.transpose(-2, -1)) / math.sqrt(d_k)
scores = scores.masked_fill(mask == 0, -1000000000.0)
p_attn = F.softmax(scores, dim=-1)
if dropout is not None:
p_attn = dropout(p_attn)
return torch.matmul(p_attn, value), p_attn
def forward(self, input_0, input_1, input_2, input_3):
primals_2 = self.linear_key.weight
primals_3 = self.linear_key.bias
primals_4 = self.linear_value.weight
primals_5 = self.linear_value.bias
primals_7 = self.linear_query.weight
primals_8 = self.linear_query.bias
primals_11 = self.linear_out.weight
primals_12 = self.linear_out.bias
primals_1 = input_0
primals_6 = input_1
primals_9 = input_2
primals_10 = input_3
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7, primals_8, primals_9,
primals_10, primals_11, primals_12])
return output[0]
| iamxpy/pointer_summarizer | MultiHeadedAttention | false | 6,855 | [
"Apache-2.0"
] | 1 | ebeb2ad32a45162c0da14dac0b6241b0b0d00fa0 | https://github.com/iamxpy/pointer_summarizer/tree/ebeb2ad32a45162c0da14dac0b6241b0b0d00fa0 | import math
import torch
import torch.nn as nn
import torch.nn.functional as F
class Model(nn.Module):
def __init__(self, num_head, d_model, dropout=0.1):
super().__init__()
assert d_model % num_head == 0
self.d_k = d_model // num_head
self.h = num_head
self.linear_key = nn.Linear(d_model, d_model)
self.linear_value = nn.Linear(d_model, d_model)
self.linear_query = nn.Linear(d_model, d_model)
self.linear_out = nn.Linear(d_model, d_model)
self.dropout = nn.Dropout(p=dropout)
def attention(self, query, key, value, mask, dropout=None):
d_k = query.size(-1)
scores = torch.matmul(query, key.transpose(-2, -1)) / math.sqrt(d_k)
scores = scores.masked_fill(mask == 0, -1000000000.0)
p_attn = F.softmax(scores, dim=-1)
if dropout is not None:
p_attn = dropout(p_attn)
return torch.matmul(p_attn, value), p_attn
def forward(self, query, key, value, mask):
nbatches = query.size(0)
query = self.linear_query(query).view(nbatches, -1, self.h, self.d_k
).transpose(1, 2)
key = self.linear_key(key).view(nbatches, -1, self.h, self.d_k
).transpose(1, 2)
value = self.linear_value(value).view(nbatches, -1, self.h, self.d_k
).transpose(1, 2)
mask = mask.unsqueeze(1)
x, _attn = self.attention(query, key, value, mask, dropout=self.dropout
)
x = x.transpose(1, 2).contiguous().view(nbatches, -1, self.h * self.d_k
)
return self.linear_out(x)
def get_inputs():
return [torch.rand([4, 4, 4]), torch.rand([4, 4, 4]), torch.rand([4, 4,
4]), torch.rand([4, 4, 4])]
def get_init_inputs():
return [4, 4]
|
A2CCritic | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/v7/cv7zazascu4rpkkwoxbiwk6c2le2e6wshdhae73bmaoapelvwguv.py
# Topologically Sorted Source Nodes: [v], Original ATen: [aten.relu, aten.threshold_backward]
# Source node to ATen node mapping:
# v => relu
# Graph fragment:
# %relu : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%view_1,), kwargs = {})
# %le_1 : [num_users=1] = call_function[target=torch.ops.aten.le.Scalar](args = (%relu, 0), kwargs = {})
triton_poi_fused_relu_threshold_backward_0 = async_compile.triton('triton_poi_fused_relu_threshold_backward_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[1024],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*i1', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_relu_threshold_backward_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_relu_threshold_backward_0(in_out_ptr0, in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 1024
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 16
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(in_out_ptr0 + (x2), tmp4, xmask)
tl.store(out_ptr0 + (x2), tmp6, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7 = args
args.clear()
assert_size_stride(primals_1, (16, 4), (4, 1))
assert_size_stride(primals_2, (16, ), (1, ))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_4, (16, 16), (16, 1))
assert_size_stride(primals_5, (16, ), (1, ))
assert_size_stride(primals_6, (1, 16), (16, 1))
assert_size_stride(primals_7, (1, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 16), (16, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(primals_3, (64, 4), (4, 1), 0), reinterpret_tensor(primals_1, (4, 16), (1, 4), 0), out=buf0)
del primals_1
buf1 = reinterpret_tensor(buf0, (4, 4, 4, 16), (256, 64, 16, 1), 0); del buf0 # reuse
buf7 = empty_strided_cuda((4, 4, 4, 16), (256, 64, 16, 1), torch.bool)
# Topologically Sorted Source Nodes: [v], Original ATen: [aten.relu, aten.threshold_backward]
stream0 = get_raw_stream(0)
triton_poi_fused_relu_threshold_backward_0.run(buf1, primals_2, buf7, 1024, grid=grid(1024), stream=stream0)
del primals_2
buf2 = empty_strided_cuda((64, 16), (16, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(buf1, (64, 16), (16, 1), 0), reinterpret_tensor(primals_4, (16, 16), (1, 16), 0), out=buf2)
buf3 = reinterpret_tensor(buf2, (4, 4, 4, 16), (256, 64, 16, 1), 0); del buf2 # reuse
buf6 = empty_strided_cuda((4, 4, 4, 16), (256, 64, 16, 1), torch.bool)
# Topologically Sorted Source Nodes: [v_1], Original ATen: [aten.relu, aten.threshold_backward]
triton_poi_fused_relu_threshold_backward_0.run(buf3, primals_5, buf6, 1024, grid=grid(1024), stream=stream0)
del primals_5
buf5 = empty_strided_cuda((64, 1), (1, 1), torch.float32)
# Topologically Sorted Source Nodes: [v_2], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_7, reinterpret_tensor(buf3, (64, 16), (16, 1), 0), reinterpret_tensor(primals_6, (16, 1), (1, 16), 0), alpha=1, beta=1, out=buf5)
del primals_7
return (reinterpret_tensor(buf5, (4, 4, 4, 1), (16, 4, 1, 1), 0), reinterpret_tensor(primals_3, (64, 4), (4, 1), 0), reinterpret_tensor(buf1, (64, 16), (16, 1), 0), reinterpret_tensor(buf3, (64, 16), (16, 1), 0), primals_6, buf6, primals_4, buf7, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((16, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((16, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((16, 16), (16, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((16, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((1, 16), (16, 1), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((1, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch as t
import torch.nn as nn
class A2CCritic(nn.Module):
def __init__(self, state_dim):
super().__init__()
self.fc1 = nn.Linear(state_dim, 16)
self.fc2 = nn.Linear(16, 16)
self.fc3 = nn.Linear(16, 1)
def forward(self, state):
v = t.relu(self.fc1(state))
v = t.relu(self.fc2(v))
v = self.fc3(v)
return v
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'state_dim': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_relu_threshold_backward_0(in_out_ptr0, in_ptr0,
out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 1024
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 16
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(in_out_ptr0 + x2, tmp4, xmask)
tl.store(out_ptr0 + x2, tmp6, xmask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7) = args
args.clear()
assert_size_stride(primals_1, (16, 4), (4, 1))
assert_size_stride(primals_2, (16,), (1,))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_4, (16, 16), (16, 1))
assert_size_stride(primals_5, (16,), (1,))
assert_size_stride(primals_6, (1, 16), (16, 1))
assert_size_stride(primals_7, (1,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 16), (16, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_3, (64, 4), (4, 1), 0),
reinterpret_tensor(primals_1, (4, 16), (1, 4), 0), out=buf0)
del primals_1
buf1 = reinterpret_tensor(buf0, (4, 4, 4, 16), (256, 64, 16, 1), 0)
del buf0
buf7 = empty_strided_cuda((4, 4, 4, 16), (256, 64, 16, 1), torch.bool)
get_raw_stream(0)
triton_poi_fused_relu_threshold_backward_0[grid(1024)](buf1,
primals_2, buf7, 1024, XBLOCK=128, num_warps=4, num_stages=1)
del primals_2
buf2 = empty_strided_cuda((64, 16), (16, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(buf1, (64, 16), (16, 1), 0),
reinterpret_tensor(primals_4, (16, 16), (1, 16), 0), out=buf2)
buf3 = reinterpret_tensor(buf2, (4, 4, 4, 16), (256, 64, 16, 1), 0)
del buf2
buf6 = empty_strided_cuda((4, 4, 4, 16), (256, 64, 16, 1), torch.bool)
triton_poi_fused_relu_threshold_backward_0[grid(1024)](buf3,
primals_5, buf6, 1024, XBLOCK=128, num_warps=4, num_stages=1)
del primals_5
buf5 = empty_strided_cuda((64, 1), (1, 1), torch.float32)
extern_kernels.addmm(primals_7, reinterpret_tensor(buf3, (64, 16),
(16, 1), 0), reinterpret_tensor(primals_6, (16, 1), (1, 16), 0),
alpha=1, beta=1, out=buf5)
del primals_7
return reinterpret_tensor(buf5, (4, 4, 4, 1), (16, 4, 1, 1), 0
), reinterpret_tensor(primals_3, (64, 4), (4, 1), 0
), reinterpret_tensor(buf1, (64, 16), (16, 1), 0), reinterpret_tensor(
buf3, (64, 16), (16, 1), 0), primals_6, buf6, primals_4, buf7
class A2CCriticNew(nn.Module):
def __init__(self, state_dim):
super().__init__()
self.fc1 = nn.Linear(state_dim, 16)
self.fc2 = nn.Linear(16, 16)
self.fc3 = nn.Linear(16, 1)
def forward(self, input_0):
primals_1 = self.fc1.weight
primals_2 = self.fc1.bias
primals_4 = self.fc2.weight
primals_5 = self.fc2.bias
primals_6 = self.fc3.weight
primals_7 = self.fc3.bias
primals_3 = input_0
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7])
return output[0]
| ikamensh/machin | A2CCritic | false | 6,856 | [
"MIT"
] | 1 | af7b423c47bc1412530cf6c96c11bd3af9b3e239 | https://github.com/ikamensh/machin/tree/af7b423c47bc1412530cf6c96c11bd3af9b3e239 | import torch
import torch as t
import torch.nn as nn
class Model(nn.Module):
def __init__(self, state_dim):
super().__init__()
self.fc1 = nn.Linear(state_dim, 16)
self.fc2 = nn.Linear(16, 16)
self.fc3 = nn.Linear(16, 1)
def forward(self, state):
v = t.relu(self.fc1(state))
v = t.relu(self.fc2(v))
v = self.fc3(v)
return v
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4]
|
DDPGCritic | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/ms/cmsuzohbg5nq52jnvirovzkvykrzzko5xomu7zyu5e5u2lhegppw.py
# Topologically Sorted Source Nodes: [state_action], Original ATen: [aten.cat]
# Source node to ATen node mapping:
# state_action => cat
# Graph fragment:
# %cat : [num_users=2] = call_function[target=torch.ops.aten.cat.default](args = ([%primals_1, %primals_2], 1), kwargs = {})
triton_poi_fused_cat_0 = async_compile.triton('triton_poi_fused_cat_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[32],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_cat_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_cat_0(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 32
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 8
x1 = (xindex // 8)
x2 = xindex
tmp0 = x0
tmp1 = tl.full([1], 0, tl.int64)
tmp2 = tmp0 >= tmp1
tmp3 = tl.full([1], 4, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = tl.load(in_ptr0 + ((4*x1) + x0), tmp4 & xmask, eviction_policy='evict_last', other=0.0)
tmp6 = tmp0 >= tmp3
tmp7 = tl.full([1], 8, tl.int64)
tmp8 = tmp0 < tmp7
tmp9 = tl.load(in_ptr1 + ((4*x1) + ((-4) + x0)), tmp6 & xmask, eviction_policy='evict_last', other=0.0)
tmp10 = tl.where(tmp4, tmp5, tmp9)
tl.store(out_ptr0 + (x2), tmp10, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/l4/cl4b2qmegciihafkh7hzuy52pdx4j53yd4lmklixpticbzk2nyq2.py
# Topologically Sorted Source Nodes: [q], Original ATen: [aten.relu]
# Source node to ATen node mapping:
# q => relu
# Graph fragment:
# %add_tensor_1 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mm_default_1, %primals_4), kwargs = {})
# %relu : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%add_tensor_1,), kwargs = {})
triton_poi_fused_relu_1 = async_compile.triton('triton_poi_fused_relu_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_relu_1', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_relu_1(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 16
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tl.store(in_out_ptr0 + (x2), tmp4, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8 = args
args.clear()
assert_size_stride(primals_1, (4, 4), (4, 1))
assert_size_stride(primals_2, (4, 4), (4, 1))
assert_size_stride(primals_3, (16, 8), (8, 1))
assert_size_stride(primals_4, (16, ), (1, ))
assert_size_stride(primals_5, (16, 16), (16, 1))
assert_size_stride(primals_6, (16, ), (1, ))
assert_size_stride(primals_7, (1, 16), (16, 1))
assert_size_stride(primals_8, (1, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 8), (8, 1), torch.float32)
# Topologically Sorted Source Nodes: [state_action], Original ATen: [aten.cat]
stream0 = get_raw_stream(0)
triton_poi_fused_cat_0.run(primals_1, primals_2, buf0, 32, grid=grid(32), stream=stream0)
del primals_1
del primals_2
buf1 = empty_strided_cuda((4, 16), (16, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(buf0, reinterpret_tensor(primals_3, (8, 16), (1, 8), 0), out=buf1)
del primals_3
buf2 = buf1; del buf1 # reuse
# Topologically Sorted Source Nodes: [q], Original ATen: [aten.relu]
triton_poi_fused_relu_1.run(buf2, primals_4, 64, grid=grid(64), stream=stream0)
del primals_4
buf3 = empty_strided_cuda((4, 16), (16, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(buf2, reinterpret_tensor(primals_5, (16, 16), (1, 16), 0), out=buf3)
buf4 = buf3; del buf3 # reuse
# Topologically Sorted Source Nodes: [q_1], Original ATen: [aten.relu]
triton_poi_fused_relu_1.run(buf4, primals_6, 64, grid=grid(64), stream=stream0)
del primals_6
buf6 = empty_strided_cuda((4, 1), (1, 1), torch.float32)
# Topologically Sorted Source Nodes: [q_2], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_8, buf4, reinterpret_tensor(primals_7, (16, 1), (1, 16), 0), alpha=1, beta=1, out=buf6)
del primals_8
return (buf6, buf0, buf2, buf4, primals_7, primals_5, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((16, 8), (8, 1), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((16, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((16, 16), (16, 1), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((16, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((1, 16), (16, 1), device='cuda:0', dtype=torch.float32)
primals_8 = rand_strided((1, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch as t
import torch.nn as nn
class DDPGCritic(nn.Module):
def __init__(self, state_dim, action_dim):
super().__init__()
self.fc1 = nn.Linear(state_dim + action_dim, 16)
self.fc2 = nn.Linear(16, 16)
self.fc3 = nn.Linear(16, 1)
def forward(self, state, action):
state_action = t.cat([state, action], 1)
q = t.relu(self.fc1(state_action))
q = t.relu(self.fc2(q))
q = self.fc3(q)
return q
def get_inputs():
return [torch.rand([4, 4]), torch.rand([4, 4])]
def get_init_inputs():
return [[], {'state_dim': 4, 'action_dim': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_cat_0(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK: tl.
constexpr):
xnumel = 32
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 8
x1 = xindex // 8
x2 = xindex
tmp0 = x0
tl.full([1], 0, tl.int64)
tmp3 = tl.full([1], 4, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = tl.load(in_ptr0 + (4 * x1 + x0), tmp4 & xmask, eviction_policy=
'evict_last', other=0.0)
tmp6 = tmp0 >= tmp3
tl.full([1], 8, tl.int64)
tmp9 = tl.load(in_ptr1 + (4 * x1 + (-4 + x0)), tmp6 & xmask,
eviction_policy='evict_last', other=0.0)
tmp10 = tl.where(tmp4, tmp5, tmp9)
tl.store(out_ptr0 + x2, tmp10, xmask)
@triton.jit
def triton_poi_fused_relu_1(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 16
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tl.store(in_out_ptr0 + x2, tmp4, xmask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7, primals_8) = args
args.clear()
assert_size_stride(primals_1, (4, 4), (4, 1))
assert_size_stride(primals_2, (4, 4), (4, 1))
assert_size_stride(primals_3, (16, 8), (8, 1))
assert_size_stride(primals_4, (16,), (1,))
assert_size_stride(primals_5, (16, 16), (16, 1))
assert_size_stride(primals_6, (16,), (1,))
assert_size_stride(primals_7, (1, 16), (16, 1))
assert_size_stride(primals_8, (1,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 8), (8, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_cat_0[grid(32)](primals_1, primals_2, buf0, 32,
XBLOCK=32, num_warps=1, num_stages=1)
del primals_1
del primals_2
buf1 = empty_strided_cuda((4, 16), (16, 1), torch.float32)
extern_kernels.mm(buf0, reinterpret_tensor(primals_3, (8, 16), (1,
8), 0), out=buf1)
del primals_3
buf2 = buf1
del buf1
triton_poi_fused_relu_1[grid(64)](buf2, primals_4, 64, XBLOCK=64,
num_warps=1, num_stages=1)
del primals_4
buf3 = empty_strided_cuda((4, 16), (16, 1), torch.float32)
extern_kernels.mm(buf2, reinterpret_tensor(primals_5, (16, 16), (1,
16), 0), out=buf3)
buf4 = buf3
del buf3
triton_poi_fused_relu_1[grid(64)](buf4, primals_6, 64, XBLOCK=64,
num_warps=1, num_stages=1)
del primals_6
buf6 = empty_strided_cuda((4, 1), (1, 1), torch.float32)
extern_kernels.addmm(primals_8, buf4, reinterpret_tensor(primals_7,
(16, 1), (1, 16), 0), alpha=1, beta=1, out=buf6)
del primals_8
return buf6, buf0, buf2, buf4, primals_7, primals_5
class DDPGCriticNew(nn.Module):
def __init__(self, state_dim, action_dim):
super().__init__()
self.fc1 = nn.Linear(state_dim + action_dim, 16)
self.fc2 = nn.Linear(16, 16)
self.fc3 = nn.Linear(16, 1)
def forward(self, input_0, input_1):
primals_3 = self.fc1.weight
primals_4 = self.fc1.bias
primals_5 = self.fc2.weight
primals_6 = self.fc2.bias
primals_7 = self.fc3.weight
primals_8 = self.fc3.bias
primals_1 = input_0
primals_2 = input_1
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7, primals_8])
return output[0]
| ikamensh/machin | DDPGCritic | false | 6,857 | [
"MIT"
] | 1 | af7b423c47bc1412530cf6c96c11bd3af9b3e239 | https://github.com/ikamensh/machin/tree/af7b423c47bc1412530cf6c96c11bd3af9b3e239 | import torch
import torch as t
import torch.nn as nn
class Model(nn.Module):
def __init__(self, state_dim, action_dim):
super().__init__()
self.fc1 = nn.Linear(state_dim + action_dim, 16)
self.fc2 = nn.Linear(16, 16)
self.fc3 = nn.Linear(16, 1)
def forward(self, state, action):
state_action = t.cat([state, action], 1)
q = t.relu(self.fc1(state_action))
q = t.relu(self.fc2(q))
q = self.fc3(q)
return q
def get_inputs():
return [torch.rand([4, 4]), torch.rand([4, 4])]
def get_init_inputs():
return [4, 4]
|
CO_Attention | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/r4/cr4mjkxmtwkqlmnjqs45p5bcvjtar5wcslrilh72fgdzrsjujqb4.py
# Topologically Sorted Source Nodes: [leaky_relu], Original ATen: [aten.leaky_relu]
# Source node to ATen node mapping:
# leaky_relu => gt, mul, where
# Graph fragment:
# %gt : [num_users=1] = call_function[target=torch.ops.aten.gt.Scalar](args = (%bmm, 0), kwargs = {})
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%bmm, 0.01), kwargs = {})
# %where : [num_users=1] = call_function[target=torch.ops.aten.where.self](args = (%gt, %bmm, %mul), kwargs = {})
triton_poi_fused_leaky_relu_0 = async_compile.triton('triton_poi_fused_leaky_relu_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_leaky_relu_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_leaky_relu_0(in_out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_out_ptr0 + (x0), xmask)
tmp1 = 0.0
tmp2 = tmp0 > tmp1
tmp3 = 0.01
tmp4 = tmp0 * tmp3
tmp5 = tl.where(tmp2, tmp0, tmp4)
tl.store(in_out_ptr0 + (x0), tmp5, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/tb/ctbtsfb24evs626gz6l7a57n4dqznpxykmjeohp2c326tad3vclf.py
# Topologically Sorted Source Nodes: [input_1, input_2], Original ATen: [aten.convolution, aten.relu]
# Source node to ATen node mapping:
# input_1 => convolution
# input_2 => relu
# Graph fragment:
# %convolution : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%unsqueeze, %primals_5, %primals_6, [1, 1], [1, 1], [1, 1], False, [0, 0], 1), kwargs = {})
# %relu : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%convolution,), kwargs = {})
triton_poi_fused_convolution_relu_1 = async_compile.triton('triton_poi_fused_convolution_relu_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[1024],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_relu_1', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_relu_1(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 1024
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 16)
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x1), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tl.store(in_out_ptr0 + (x2), tmp4, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/xj/cxjegmugqxkfvjsmgyyjy6pudainaptuugpnqcqjyfhnho3vnzf6.py
# Topologically Sorted Source Nodes: [input_3], Original ATen: [aten.max_pool2d_with_indices]
# Source node to ATen node mapping:
# input_3 => getitem, getitem_1
# Graph fragment:
# %getitem : [num_users=2] = call_function[target=operator.getitem](args = (%_low_memory_max_pool2d_with_offsets, 0), kwargs = {})
# %getitem_1 : [num_users=1] = call_function[target=operator.getitem](args = (%_low_memory_max_pool2d_with_offsets, 1), kwargs = {})
triton_poi_fused_max_pool2d_with_indices_2 = async_compile.triton('triton_poi_fused_max_pool2d_with_indices_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*i8', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_max_pool2d_with_indices_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 4, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_max_pool2d_with_indices_2(in_ptr0, out_ptr0, out_ptr1, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 2
x1 = (xindex // 2)
x2 = xindex
tmp0 = tl.load(in_ptr0 + ((2*x0) + (8*x1)), xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr0 + (1 + (2*x0) + (8*x1)), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr0 + (4 + (2*x0) + (8*x1)), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (5 + (2*x0) + (8*x1)), xmask, eviction_policy='evict_last')
tmp2 = triton_helpers.maximum(tmp1, tmp0)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp6 = triton_helpers.maximum(tmp5, tmp4)
tmp7 = tmp1 > tmp0
tmp8 = tl.full([1], 1, tl.int8)
tmp9 = tl.full([1], 0, tl.int8)
tmp10 = tl.where(tmp7, tmp8, tmp9)
tmp11 = tmp3 > tmp2
tmp12 = tl.full([1], 2, tl.int8)
tmp13 = tl.where(tmp11, tmp12, tmp10)
tmp14 = tmp5 > tmp4
tmp15 = tl.full([1], 3, tl.int8)
tmp16 = tl.where(tmp14, tmp15, tmp13)
tl.store(out_ptr0 + (x2), tmp6, xmask)
tl.store(out_ptr1 + (x2), tmp16, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/du/cduvnnmoknjnz3urwik3lbda334jqqxr5nfhefejobdpe7rtqosd.py
# Topologically Sorted Source Nodes: [input_4, input_5], Original ATen: [aten.convolution, aten.relu, aten.threshold_backward]
# Source node to ATen node mapping:
# input_4 => convolution_1
# input_5 => relu_1
# Graph fragment:
# %convolution_1 : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%getitem, %primals_7, %primals_8, [1, 1], [1, 1], [1, 1], False, [0, 0], 1), kwargs = {})
# %relu_1 : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%convolution_1,), kwargs = {})
# %le_2 : [num_users=1] = call_function[target=torch.ops.aten.le.Scalar](args = (%relu_1, 0), kwargs = {})
triton_poi_fused_convolution_relu_threshold_backward_3 = async_compile.triton('triton_poi_fused_convolution_relu_threshold_backward_3', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*i1', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_relu_threshold_backward_3', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_relu_threshold_backward_3(in_out_ptr0, in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x1), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(in_out_ptr0 + (x2), tmp4, xmask)
tl.store(out_ptr0 + (x2), tmp6, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/bq/cbqsq6ivu2wn7dt6dhn4vyu6ieqsx2wlvlk7e5paqjugqvkxyxl7.py
# Topologically Sorted Source Nodes: [attention_1], Original ATen: [aten._to_copy]
# Source node to ATen node mapping:
# attention_1 => convert_element_type_1
# Graph fragment:
# %convert_element_type_1 : [num_users=5] = call_function[target=torch.ops.prims.convert_element_type.default](args = (%view_6, torch.int64), kwargs = {})
triton_poi_fused__to_copy_4 = async_compile.triton('triton_poi_fused__to_copy_4', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*i64', 1: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__to_copy_4', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 0, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__to_copy_4(out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = x0
tmp1 = tmp0.to(tl.float32)
tmp2 = 0.015873015873015872
tmp3 = tmp1 * tmp2
tmp4 = 0.0
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp6 = tmp5.to(tl.int32)
tl.store(out_ptr0 + (x0), tmp6, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/id/cidz4bm5dmsgdfl4wf24t3hw6ypb6nrmq3jkygguxra5qhktjdch.py
# Topologically Sorted Source Nodes: [attention_1], Original ATen: [aten.add, aten.clamp]
# Source node to ATen node mapping:
# attention_1 => add_1, clamp_max
# Graph fragment:
# %add_1 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%convert_element_type_1, 1), kwargs = {})
# %clamp_max : [num_users=3] = call_function[target=torch.ops.aten.clamp_max.default](args = (%add_1, 1), kwargs = {})
triton_poi_fused_add_clamp_5 = async_compile.triton('triton_poi_fused_add_clamp_5', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*i64', 1: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_clamp_5', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 0, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_clamp_5(out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = x0
tmp1 = tmp0.to(tl.float32)
tmp2 = 0.015873015873015872
tmp3 = tmp1 * tmp2
tmp4 = 0.0
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp6 = tmp5.to(tl.int32)
tmp7 = tl.full([1], 1, tl.int64)
tmp8 = tmp6 + tmp7
tmp9 = triton_helpers.minimum(tmp8, tmp7)
tl.store(out_ptr0 + (x0), tmp9, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/s2/cs2bfl3szvqhjsn2mvzxz4k474uvcr67f6v2dslkqak5yqd4xmdf.py
# Topologically Sorted Source Nodes: [attention_1], Original ATen: [aten.arange, aten._to_copy, aten.mul, aten.clamp, aten.sub]
# Source node to ATen node mapping:
# attention_1 => clamp_max_2, clamp_min, clamp_min_2, convert_element_type, iota, mul_2, sub
# Graph fragment:
# %iota : [num_users=1] = call_function[target=torch.ops.prims.iota.default](args = (64,), kwargs = {start: 0, step: 1, dtype: torch.int64, device: cuda:0, requires_grad: False})
# %convert_element_type : [num_users=1] = call_function[target=torch.ops.prims.convert_element_type.default](args = (%iota, torch.float32), kwargs = {})
# %mul_2 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%convert_element_type, 0.015873015873015872), kwargs = {})
# %clamp_min : [num_users=3] = call_function[target=torch.ops.aten.clamp_min.default](args = (%mul_2, 0.0), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%clamp_min, %convert_element_type_3), kwargs = {})
# %clamp_min_2 : [num_users=1] = call_function[target=torch.ops.aten.clamp_min.default](args = (%sub, 0.0), kwargs = {})
# %clamp_max_2 : [num_users=3] = call_function[target=torch.ops.aten.clamp_max.default](args = (%clamp_min_2, 1.0), kwargs = {})
triton_poi_fused__to_copy_arange_clamp_mul_sub_6 = async_compile.triton('triton_poi_fused__to_copy_arange_clamp_mul_sub_6', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__to_copy_arange_clamp_mul_sub_6', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 0, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__to_copy_arange_clamp_mul_sub_6(out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = x0
tmp1 = tmp0.to(tl.float32)
tmp2 = 0.015873015873015872
tmp3 = tmp1 * tmp2
tmp4 = 0.0
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp6 = tmp5.to(tl.int32)
tmp7 = tmp6.to(tl.float32)
tmp8 = tmp5 - tmp7
tmp9 = triton_helpers.maximum(tmp8, tmp4)
tmp10 = 1.0
tmp11 = triton_helpers.minimum(tmp9, tmp10)
tl.store(out_ptr0 + (x0), tmp11, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/xq/cxqyzvjx3e2c65teved7fjysjg5bcwvmoatvep23verfhqd4jt4h.py
# Topologically Sorted Source Nodes: [attention_1], Original ATen: [aten._unsafe_index, aten.sub, aten.mul, aten.add]
# Source node to ATen node mapping:
# attention_1 => _unsafe_index, _unsafe_index_1, _unsafe_index_2, _unsafe_index_3, add_3, add_4, add_5, mul_4, mul_5, mul_6, sub_1, sub_2, sub_4
# Graph fragment:
# %_unsafe_index : [num_users=2] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%view_5, [None, None, %convert_element_type_1, %convert_element_type_3]), kwargs = {})
# %_unsafe_index_1 : [num_users=1] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%view_5, [None, None, %convert_element_type_1, %clamp_max_1]), kwargs = {})
# %_unsafe_index_2 : [num_users=2] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%view_5, [None, None, %clamp_max, %convert_element_type_3]), kwargs = {})
# %_unsafe_index_3 : [num_users=1] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%view_5, [None, None, %clamp_max, %clamp_max_1]), kwargs = {})
# %sub_1 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%_unsafe_index_1, %_unsafe_index), kwargs = {})
# %mul_4 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sub_1, %clamp_max_2), kwargs = {})
# %add_3 : [num_users=2] = call_function[target=torch.ops.aten.add.Tensor](args = (%_unsafe_index, %mul_4), kwargs = {})
# %sub_2 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%_unsafe_index_3, %_unsafe_index_2), kwargs = {})
# %mul_5 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sub_2, %clamp_max_2), kwargs = {})
# %add_4 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%_unsafe_index_2, %mul_5), kwargs = {})
# %sub_4 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%add_4, %add_3), kwargs = {})
# %mul_6 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sub_4, %clamp_max_3), kwargs = {})
# %add_5 : [num_users=2] = call_function[target=torch.ops.aten.add.Tensor](args = (%add_3, %mul_6), kwargs = {})
triton_poi_fused__unsafe_index_add_mul_sub_7 = async_compile.triton('triton_poi_fused__unsafe_index_add_mul_sub_7', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[4096],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*i64', 2: '*i64', 3: '*fp32', 4: '*fp32', 5: '*i64', 6: '*fp32', 7: '*i64', 8: '*fp32', 9: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5, 6, 7, 8, 9), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__unsafe_index_add_mul_sub_7', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 7, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__unsafe_index_add_mul_sub_7(in_out_ptr0, in_ptr0, in_ptr1, in_ptr2, in_ptr3, in_ptr4, in_ptr5, in_ptr6, in_ptr7, xnumel, XBLOCK : tl.constexpr):
xnumel = 4096
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = tl.full([XBLOCK], True, tl.int1)
x1 = (xindex // 64)
x0 = xindex % 64
x2 = xindex
tmp0 = tl.load(in_ptr0 + (x1), None, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr1 + (x0), None, eviction_policy='evict_last')
tmp10 = tl.load(in_ptr3 + (0))
tmp11 = tl.broadcast_to(tmp10, [XBLOCK])
tmp13 = tl.load(in_ptr4 + (x0), None, eviction_policy='evict_last')
tmp20 = tl.load(in_ptr5 + (x0), None, eviction_policy='evict_last')
tmp23 = tl.load(in_ptr6 + (x1), None, eviction_policy='evict_last')
tmp35 = tl.load(in_ptr7 + (x1), None, eviction_policy='evict_last')
tmp1 = tl.full([XBLOCK], 2, tl.int32)
tmp2 = tmp0 + tmp1
tmp3 = tmp0 < 0
tmp4 = tl.where(tmp3, tmp2, tmp0)
tmp6 = tmp5 + tmp1
tmp7 = tmp5 < 0
tmp8 = tl.where(tmp7, tmp6, tmp5)
tmp9 = tl.load(in_ptr2 + (tmp8 + (2*tmp4)), None, eviction_policy='evict_last')
tmp12 = tmp9 + tmp11
tmp14 = tmp13 + tmp1
tmp15 = tmp13 < 0
tmp16 = tl.where(tmp15, tmp14, tmp13)
tmp17 = tl.load(in_ptr2 + (tmp16 + (2*tmp4)), None, eviction_policy='evict_last')
tmp18 = tmp17 + tmp11
tmp19 = tmp18 - tmp12
tmp21 = tmp19 * tmp20
tmp22 = tmp12 + tmp21
tmp24 = tmp23 + tmp1
tmp25 = tmp23 < 0
tmp26 = tl.where(tmp25, tmp24, tmp23)
tmp27 = tl.load(in_ptr2 + (tmp8 + (2*tmp26)), None, eviction_policy='evict_last')
tmp28 = tmp27 + tmp11
tmp29 = tl.load(in_ptr2 + (tmp16 + (2*tmp26)), None, eviction_policy='evict_last')
tmp30 = tmp29 + tmp11
tmp31 = tmp30 - tmp28
tmp32 = tmp31 * tmp20
tmp33 = tmp28 + tmp32
tmp34 = tmp33 - tmp22
tmp36 = tmp34 * tmp35
tmp37 = tmp22 + tmp36
tl.store(in_out_ptr0 + (x2), tmp37, None)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/e4/ce4yoq25u46wevqfnz37rbjcwdllazvb624bgh7gdwfnaytstgeq.py
# Topologically Sorted Source Nodes: [interpolate_1], Original ATen: [aten._to_copy]
# Source node to ATen node mapping:
# interpolate_1 => convert_element_type_5
# Graph fragment:
# %convert_element_type_5 : [num_users=5] = call_function[target=torch.ops.prims.convert_element_type.default](args = (%view_8, torch.int64), kwargs = {})
triton_poi_fused__to_copy_8 = async_compile.triton('triton_poi_fused__to_copy_8', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[4],
filename=__file__,
triton_meta={'signature': {0: '*i64', 1: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0,), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__to_copy_8', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 0, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__to_copy_8(out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 4
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = x0
tmp1 = tmp0.to(tl.float32)
tmp2 = 21.0
tmp3 = tmp1 * tmp2
tmp4 = 0.0
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp6 = tmp5.to(tl.int32)
tl.store(out_ptr0 + (x0), tmp6, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/fk/cfkzacirmz5n2iflq3uev5aqmta5rxnbevjxabubygs2p7s73pmh.py
# Topologically Sorted Source Nodes: [interpolate_1], Original ATen: [aten.add, aten.clamp]
# Source node to ATen node mapping:
# interpolate_1 => add_6, clamp_max_4
# Graph fragment:
# %add_6 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%convert_element_type_5, 1), kwargs = {})
# %clamp_max_4 : [num_users=3] = call_function[target=torch.ops.aten.clamp_max.default](args = (%add_6, 63), kwargs = {})
triton_poi_fused_add_clamp_9 = async_compile.triton('triton_poi_fused_add_clamp_9', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[4],
filename=__file__,
triton_meta={'signature': {0: '*i64', 1: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0,), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_clamp_9', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 0, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_clamp_9(out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 4
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = x0
tmp1 = tmp0.to(tl.float32)
tmp2 = 21.0
tmp3 = tmp1 * tmp2
tmp4 = 0.0
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp6 = tmp5.to(tl.int32)
tmp7 = tl.full([1], 1, tl.int64)
tmp8 = tmp6 + tmp7
tmp9 = tl.full([1], 63, tl.int64)
tmp10 = triton_helpers.minimum(tmp8, tmp9)
tl.store(out_ptr0 + (x0), tmp10, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/hs/chs5sft3pj55v3xkxmw5em3r3jnrjbyozjvo42gtczmnep5ilzqs.py
# Topologically Sorted Source Nodes: [interpolate_1], Original ATen: [aten.arange, aten._to_copy, aten.mul, aten.clamp, aten.sub]
# Source node to ATen node mapping:
# interpolate_1 => clamp_max_6, clamp_min_4, clamp_min_6, convert_element_type_4, iota_2, mul_7, sub_5
# Graph fragment:
# %iota_2 : [num_users=1] = call_function[target=torch.ops.prims.iota.default](args = (4,), kwargs = {start: 0, step: 1, dtype: torch.int64, device: cuda:0, requires_grad: False})
# %convert_element_type_4 : [num_users=1] = call_function[target=torch.ops.prims.convert_element_type.default](args = (%iota_2, torch.float32), kwargs = {})
# %mul_7 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%convert_element_type_4, 21.0), kwargs = {})
# %clamp_min_4 : [num_users=3] = call_function[target=torch.ops.aten.clamp_min.default](args = (%mul_7, 0.0), kwargs = {})
# %sub_5 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%clamp_min_4, %convert_element_type_7), kwargs = {})
# %clamp_min_6 : [num_users=1] = call_function[target=torch.ops.aten.clamp_min.default](args = (%sub_5, 0.0), kwargs = {})
# %clamp_max_6 : [num_users=3] = call_function[target=torch.ops.aten.clamp_max.default](args = (%clamp_min_6, 1.0), kwargs = {})
triton_poi_fused__to_copy_arange_clamp_mul_sub_10 = async_compile.triton('triton_poi_fused__to_copy_arange_clamp_mul_sub_10', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[4],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0,), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__to_copy_arange_clamp_mul_sub_10', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 0, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__to_copy_arange_clamp_mul_sub_10(out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 4
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = x0
tmp1 = tmp0.to(tl.float32)
tmp2 = 21.0
tmp3 = tmp1 * tmp2
tmp4 = 0.0
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp6 = tmp5.to(tl.int32)
tmp7 = tmp6.to(tl.float32)
tmp8 = tmp5 - tmp7
tmp9 = triton_helpers.maximum(tmp8, tmp4)
tmp10 = 1.0
tmp11 = triton_helpers.minimum(tmp9, tmp10)
tl.store(out_ptr0 + (x0), tmp11, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/e2/ce2cztqnpna3ht6cnb4dfnbpe2sy4fp3cm2rtlexrpj3aomh2kob.py
# Topologically Sorted Source Nodes: [attention_2, interpolate_1], Original ATen: [aten.convolution, aten._unsafe_index, aten.sub, aten.mul, aten.add]
# Source node to ATen node mapping:
# attention_2 => convolution_4
# interpolate_1 => _unsafe_index_4, _unsafe_index_5, _unsafe_index_6, _unsafe_index_7, add_8, add_9, mul_10, mul_11, mul_9, sub_6, sub_7, sub_9
# Graph fragment:
# %convolution_4 : [num_users=4] = call_function[target=torch.ops.aten.convolution.default](args = (%add_5, %primals_15, %primals_16, [1, 1], [1, 1], [1, 1], False, [0, 0], 1), kwargs = {})
# %_unsafe_index_4 : [num_users=2] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%convolution_4, [None, None, %convert_element_type_5, %convert_element_type_7]), kwargs = {})
# %_unsafe_index_5 : [num_users=1] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%convolution_4, [None, None, %convert_element_type_5, %clamp_max_5]), kwargs = {})
# %_unsafe_index_6 : [num_users=2] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%convolution_4, [None, None, %clamp_max_4, %convert_element_type_7]), kwargs = {})
# %_unsafe_index_7 : [num_users=1] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%convolution_4, [None, None, %clamp_max_4, %clamp_max_5]), kwargs = {})
# %sub_6 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%_unsafe_index_5, %_unsafe_index_4), kwargs = {})
# %mul_9 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sub_6, %clamp_max_6), kwargs = {})
# %add_8 : [num_users=2] = call_function[target=torch.ops.aten.add.Tensor](args = (%_unsafe_index_4, %mul_9), kwargs = {})
# %sub_7 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%_unsafe_index_7, %_unsafe_index_6), kwargs = {})
# %mul_10 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sub_7, %clamp_max_6), kwargs = {})
# %add_9 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%_unsafe_index_6, %mul_10), kwargs = {})
# %sub_9 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%add_9, %add_8), kwargs = {})
# %mul_11 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sub_9, %clamp_max_7), kwargs = {})
triton_poi_fused__unsafe_index_add_convolution_mul_sub_11 = async_compile.triton('triton_poi_fused__unsafe_index_add_convolution_mul_sub_11', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*i64', 1: '*i64', 2: '*fp32', 3: '*fp32', 4: '*i64', 5: '*fp32', 6: '*i64', 7: '*fp32', 8: '*fp32', 9: '*fp32', 10: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__unsafe_index_add_convolution_mul_sub_11', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 7, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__unsafe_index_add_convolution_mul_sub_11(in_ptr0, in_ptr1, in_ptr2, in_ptr3, in_ptr4, in_ptr5, in_ptr6, in_ptr7, out_ptr0, out_ptr1, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = (xindex // 4)
x0 = xindex % 4
x2 = xindex
tmp0 = tl.load(in_ptr0 + (x1), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr1 + (x0), xmask, eviction_policy='evict_last')
tmp10 = tl.load(in_ptr3 + (0))
tmp11 = tl.broadcast_to(tmp10, [XBLOCK])
tmp13 = tl.load(in_ptr4 + (x0), xmask, eviction_policy='evict_last')
tmp20 = tl.load(in_ptr5 + (x0), xmask, eviction_policy='evict_last')
tmp23 = tl.load(in_ptr6 + (x1), xmask, eviction_policy='evict_last')
tmp35 = tl.load(in_ptr7 + (x1), xmask, eviction_policy='evict_last')
tmp1 = tl.full([XBLOCK], 64, tl.int32)
tmp2 = tmp0 + tmp1
tmp3 = tmp0 < 0
tmp4 = tl.where(tmp3, tmp2, tmp0)
tmp6 = tmp5 + tmp1
tmp7 = tmp5 < 0
tmp8 = tl.where(tmp7, tmp6, tmp5)
tmp9 = tl.load(in_ptr2 + (tmp8 + (64*tmp4)), xmask, eviction_policy='evict_last')
tmp12 = tmp9 + tmp11
tmp14 = tmp13 + tmp1
tmp15 = tmp13 < 0
tmp16 = tl.where(tmp15, tmp14, tmp13)
tmp17 = tl.load(in_ptr2 + (tmp16 + (64*tmp4)), xmask, eviction_policy='evict_last')
tmp18 = tmp17 + tmp11
tmp19 = tmp18 - tmp12
tmp21 = tmp19 * tmp20
tmp22 = tmp12 + tmp21
tmp24 = tmp23 + tmp1
tmp25 = tmp23 < 0
tmp26 = tl.where(tmp25, tmp24, tmp23)
tmp27 = tl.load(in_ptr2 + (tmp8 + (64*tmp26)), xmask, eviction_policy='evict_last')
tmp28 = tmp27 + tmp11
tmp29 = tl.load(in_ptr2 + (tmp16 + (64*tmp26)), xmask, eviction_policy='evict_last')
tmp30 = tmp29 + tmp11
tmp31 = tmp30 - tmp28
tmp32 = tmp31 * tmp20
tmp33 = tmp28 + tmp32
tmp34 = tmp33 - tmp22
tmp36 = tmp34 * tmp35
tl.store(out_ptr0 + (x2), tmp22, xmask)
tl.store(out_ptr1 + (x2), tmp36, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/al/calhm7qaxirnnipgjlsqqztidolqmscv6mt7chen2lj6kba42y6y.py
# Topologically Sorted Source Nodes: [x_1], Original ATen: [aten.sigmoid]
# Source node to ATen node mapping:
# x_1 => sigmoid
# Graph fragment:
# %sigmoid : [num_users=1] = call_function[target=torch.ops.aten.sigmoid.default](args = (%permute_4,), kwargs = {})
triton_poi_fused_sigmoid_12 = async_compile.triton('triton_poi_fused_sigmoid_12', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_sigmoid_12', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 3, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_sigmoid_12(in_ptr0, in_ptr1, in_ptr2, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = (xindex // 4)
x2 = xindex
tmp0 = tl.load(in_ptr0 + (x1), xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (x1), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr2 + (x2), xmask)
tmp2 = tmp0 + tmp1
tmp4 = tmp2 * tmp3
tmp5 = tl.sigmoid(tmp4)
tl.store(out_ptr0 + (x2), tmp5, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9, primals_10, primals_11, primals_12, primals_13, primals_14, primals_15, primals_16 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_2, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_3, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_4, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_5, (64, 4, 3, 3), (36, 9, 3, 1))
assert_size_stride(primals_6, (64, ), (1, ))
assert_size_stride(primals_7, (4, 64, 3, 3), (576, 9, 3, 1))
assert_size_stride(primals_8, (4, ), (1, ))
assert_size_stride(primals_9, (64, 4, 3, 3), (36, 9, 3, 1))
assert_size_stride(primals_10, (64, ), (1, ))
assert_size_stride(primals_11, (4, 64, 3, 3), (576, 9, 3, 1))
assert_size_stride(primals_12, (4, ), (1, ))
assert_size_stride(primals_13, (1, 4, 4), (16, 4, 1))
assert_size_stride(primals_14, (1, ), (1, ))
assert_size_stride(primals_15, (1, 1, 3, 3), (9, 9, 3, 1))
assert_size_stride(primals_16, (1, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [bmm], Original ATen: [aten.bmm]
extern_kernels.bmm(reinterpret_tensor(primals_4, (4, 4, 4), (1, 4, 16), 0), reinterpret_tensor(primals_1, (4, 4, 4), (1, 16, 4), 0), out=buf0)
del primals_4
buf1 = buf0; del buf0 # reuse
# Topologically Sorted Source Nodes: [leaky_relu], Original ATen: [aten.leaky_relu]
stream0 = get_raw_stream(0)
triton_poi_fused_leaky_relu_0.run(buf1, 64, grid=grid(64), stream=stream0)
buf2 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [bmm_1], Original ATen: [aten.bmm]
extern_kernels.bmm(reinterpret_tensor(primals_3, (4, 4, 4), (1, 4, 16), 0), reinterpret_tensor(primals_1, (4, 4, 4), (1, 16, 4), 0), out=buf2)
del primals_1
del primals_3
buf3 = buf2; del buf2 # reuse
# Topologically Sorted Source Nodes: [leaky_relu_1], Original ATen: [aten.leaky_relu]
triton_poi_fused_leaky_relu_0.run(buf3, 64, grid=grid(64), stream=stream0)
# Topologically Sorted Source Nodes: [input_1], Original ATen: [aten.convolution]
buf4 = extern_kernels.convolution(reinterpret_tensor(buf1, (1, 4, 4, 4), (0, 16, 4, 1), 0), primals_5, stride=(1, 1), padding=(1, 1), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf4, (1, 64, 4, 4), (1024, 16, 4, 1))
buf5 = buf4; del buf4 # reuse
# Topologically Sorted Source Nodes: [input_1, input_2], Original ATen: [aten.convolution, aten.relu]
triton_poi_fused_convolution_relu_1.run(buf5, primals_6, 1024, grid=grid(1024), stream=stream0)
del primals_6
buf6 = empty_strided_cuda((1, 64, 2, 2), (256, 4, 2, 1), torch.float32)
buf7 = empty_strided_cuda((1, 64, 2, 2), (256, 4, 2, 1), torch.int8)
# Topologically Sorted Source Nodes: [input_3], Original ATen: [aten.max_pool2d_with_indices]
triton_poi_fused_max_pool2d_with_indices_2.run(buf5, buf6, buf7, 256, grid=grid(256), stream=stream0)
# Topologically Sorted Source Nodes: [input_4], Original ATen: [aten.convolution]
buf8 = extern_kernels.convolution(buf6, primals_7, stride=(1, 1), padding=(1, 1), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf8, (1, 4, 2, 2), (16, 4, 2, 1))
# Topologically Sorted Source Nodes: [input_6], Original ATen: [aten.convolution]
buf9 = extern_kernels.convolution(reinterpret_tensor(buf3, (1, 4, 4, 4), (0, 16, 4, 1), 0), primals_9, stride=(1, 1), padding=(1, 1), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf9, (1, 64, 4, 4), (1024, 16, 4, 1))
buf10 = buf9; del buf9 # reuse
# Topologically Sorted Source Nodes: [input_6, input_7], Original ATen: [aten.convolution, aten.relu]
triton_poi_fused_convolution_relu_1.run(buf10, primals_10, 1024, grid=grid(1024), stream=stream0)
del primals_10
buf11 = empty_strided_cuda((1, 64, 2, 2), (256, 4, 2, 1), torch.float32)
buf12 = empty_strided_cuda((1, 64, 2, 2), (256, 4, 2, 1), torch.int8)
# Topologically Sorted Source Nodes: [input_8], Original ATen: [aten.max_pool2d_with_indices]
triton_poi_fused_max_pool2d_with_indices_2.run(buf10, buf11, buf12, 256, grid=grid(256), stream=stream0)
# Topologically Sorted Source Nodes: [input_9], Original ATen: [aten.convolution]
buf13 = extern_kernels.convolution(buf11, primals_11, stride=(1, 1), padding=(1, 1), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf13, (1, 4, 2, 2), (16, 4, 2, 1))
buf14 = buf8; del buf8 # reuse
buf38 = empty_strided_cuda((1, 4, 2, 2), (16, 4, 2, 1), torch.bool)
# Topologically Sorted Source Nodes: [input_4, input_5], Original ATen: [aten.convolution, aten.relu, aten.threshold_backward]
triton_poi_fused_convolution_relu_threshold_backward_3.run(buf14, primals_8, buf38, 16, grid=grid(16), stream=stream0)
del primals_8
buf15 = buf13; del buf13 # reuse
buf37 = empty_strided_cuda((1, 4, 2, 2), (16, 4, 2, 1), torch.bool)
# Topologically Sorted Source Nodes: [input_9, input_10], Original ATen: [aten.convolution, aten.relu, aten.threshold_backward]
triton_poi_fused_convolution_relu_threshold_backward_3.run(buf15, primals_12, buf37, 16, grid=grid(16), stream=stream0)
del primals_12
# Topologically Sorted Source Nodes: [bilinear], Original ATen: [aten._trilinear]
buf16 = torch.ops.aten._trilinear.default(reinterpret_tensor(buf14, (4, 4), (4, 1), 0), primals_13, reinterpret_tensor(buf15, (4, 4), (4, 1), 0), [1, 3], [0], [1, 2], [2, 3])
buf17 = buf16
del buf16
buf18 = empty_strided_cuda((64, 1), (1, 1), torch.int64)
# Topologically Sorted Source Nodes: [attention_1], Original ATen: [aten._to_copy]
triton_poi_fused__to_copy_4.run(buf18, 64, grid=grid(64), stream=stream0)
buf19 = empty_strided_cuda((64, 1), (1, 1), torch.int64)
# Topologically Sorted Source Nodes: [attention_1], Original ATen: [aten.add, aten.clamp]
triton_poi_fused_add_clamp_5.run(buf19, 64, grid=grid(64), stream=stream0)
buf20 = empty_strided_cuda((64, ), (1, ), torch.int64)
# Topologically Sorted Source Nodes: [attention_1], Original ATen: [aten.arange, aten._to_copy, aten.mul, aten.clamp]
triton_poi_fused__to_copy_4.run(buf20, 64, grid=grid(64), stream=stream0)
buf21 = empty_strided_cuda((64, ), (1, ), torch.int64)
# Topologically Sorted Source Nodes: [attention_1], Original ATen: [aten.add, aten.clamp]
triton_poi_fused_add_clamp_5.run(buf21, 64, grid=grid(64), stream=stream0)
buf22 = empty_strided_cuda((64, ), (1, ), torch.float32)
# Topologically Sorted Source Nodes: [attention_1], Original ATen: [aten.arange, aten._to_copy, aten.mul, aten.clamp, aten.sub]
triton_poi_fused__to_copy_arange_clamp_mul_sub_6.run(buf22, 64, grid=grid(64), stream=stream0)
buf24 = empty_strided_cuda((64, 1), (1, 1), torch.float32)
# Topologically Sorted Source Nodes: [attention_1], Original ATen: [aten.sub, aten.clamp]
triton_poi_fused__to_copy_arange_clamp_mul_sub_6.run(buf24, 64, grid=grid(64), stream=stream0)
buf25 = empty_strided_cuda((1, 1, 64, 64), (4096, 4096, 64, 1), torch.float32)
buf26 = buf25; del buf25 # reuse
# Topologically Sorted Source Nodes: [attention_1], Original ATen: [aten._unsafe_index, aten.sub, aten.mul, aten.add]
triton_poi_fused__unsafe_index_add_mul_sub_7.run(buf26, buf18, buf20, buf17, primals_14, buf21, buf22, buf19, buf24, 4096, grid=grid(4096), stream=stream0)
del primals_14
# Topologically Sorted Source Nodes: [attention_2], Original ATen: [aten.convolution]
buf27 = extern_kernels.convolution(buf26, primals_15, stride=(1, 1), padding=(1, 1), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf27, (1, 1, 64, 64), (4096, 4096, 64, 1))
buf28 = empty_strided_cuda((4, 1), (1, 1), torch.int64)
# Topologically Sorted Source Nodes: [interpolate_1], Original ATen: [aten._to_copy]
triton_poi_fused__to_copy_8.run(buf28, 4, grid=grid(4), stream=stream0)
buf29 = empty_strided_cuda((4, 1), (1, 1), torch.int64)
# Topologically Sorted Source Nodes: [interpolate_1], Original ATen: [aten.add, aten.clamp]
triton_poi_fused_add_clamp_9.run(buf29, 4, grid=grid(4), stream=stream0)
buf30 = empty_strided_cuda((4, ), (1, ), torch.int64)
# Topologically Sorted Source Nodes: [interpolate_1], Original ATen: [aten.arange, aten._to_copy, aten.mul, aten.clamp]
triton_poi_fused__to_copy_8.run(buf30, 4, grid=grid(4), stream=stream0)
buf31 = empty_strided_cuda((4, ), (1, ), torch.int64)
# Topologically Sorted Source Nodes: [interpolate_1], Original ATen: [aten.add, aten.clamp]
triton_poi_fused_add_clamp_9.run(buf31, 4, grid=grid(4), stream=stream0)
buf32 = reinterpret_tensor(buf17, (4, ), (1, ), 0); del buf17 # reuse
# Topologically Sorted Source Nodes: [interpolate_1], Original ATen: [aten.arange, aten._to_copy, aten.mul, aten.clamp, aten.sub]
triton_poi_fused__to_copy_arange_clamp_mul_sub_10.run(buf32, 4, grid=grid(4), stream=stream0)
buf34 = empty_strided_cuda((4, 1), (1, 1), torch.float32)
# Topologically Sorted Source Nodes: [interpolate_1], Original ATen: [aten.sub, aten.clamp]
triton_poi_fused__to_copy_arange_clamp_mul_sub_10.run(buf34, 4, grid=grid(4), stream=stream0)
buf33 = empty_strided_cuda((1, 1, 4, 4), (16, 16, 4, 1), torch.float32)
buf35 = empty_strided_cuda((1, 1, 4, 4), (16, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [attention_2, interpolate_1], Original ATen: [aten.convolution, aten._unsafe_index, aten.sub, aten.mul, aten.add]
triton_poi_fused__unsafe_index_add_convolution_mul_sub_11.run(buf28, buf30, buf27, primals_16, buf31, buf32, buf29, buf34, buf33, buf35, 16, grid=grid(16), stream=stream0)
del buf27
del primals_16
buf36 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 1), torch.float32)
# Topologically Sorted Source Nodes: [x_1], Original ATen: [aten.sigmoid]
triton_poi_fused_sigmoid_12.run(buf33, buf35, primals_2, buf36, 64, grid=grid(64), stream=stream0)
del buf33
del buf35
return (buf36, primals_2, primals_5, primals_7, primals_9, primals_11, primals_13, primals_15, reinterpret_tensor(buf1, (1, 4, 4, 4), (64, 16, 4, 1), 0), reinterpret_tensor(buf3, (1, 4, 4, 4), (64, 16, 4, 1), 0), buf5, buf6, buf7, buf10, buf11, buf12, reinterpret_tensor(buf14, (4, 4), (4, 1), 0), reinterpret_tensor(buf15, (4, 4), (4, 1), 0), buf18, buf19, buf20, buf21, buf22, buf24, buf26, buf28, buf29, buf30, buf31, buf32, buf34, buf36, buf37, buf38, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((64, 4, 3, 3), (36, 9, 3, 1), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((64, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((4, 64, 3, 3), (576, 9, 3, 1), device='cuda:0', dtype=torch.float32)
primals_8 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_9 = rand_strided((64, 4, 3, 3), (36, 9, 3, 1), device='cuda:0', dtype=torch.float32)
primals_10 = rand_strided((64, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_11 = rand_strided((4, 64, 3, 3), (576, 9, 3, 1), device='cuda:0', dtype=torch.float32)
primals_12 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_13 = rand_strided((1, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_14 = rand_strided((1, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_15 = rand_strided((1, 1, 3, 3), (9, 9, 3, 1), device='cuda:0', dtype=torch.float32)
primals_16 = rand_strided((1, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9, primals_10, primals_11, primals_12, primals_13, primals_14, primals_15, primals_16])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.nn.functional as F
class CO_Attention(nn.Module):
def __init__(self, in_dim, co_attention_dim):
super(CO_Attention, self).__init__()
self.leak_relu = nn.LeakyReLU()
self.relu = nn.ReLU()
self.conv1 = nn.Conv2d(in_dim, 64, kernel_size=3, padding=1)
self.conv2 = nn.Conv2d(64, co_attention_dim, kernel_size=3, padding=1)
self.pool = torch.nn.MaxPool2d(kernel_size=2)
self.gate1 = nn.Sequential(self.conv1, self.relu, self.pool, self.
conv2, self.relu)
self.conv3 = nn.Conv2d(in_dim, 64, kernel_size=3, padding=1)
self.conv4 = nn.Conv2d(64, co_attention_dim, kernel_size=3, padding=1)
self.gate2 = nn.Sequential(self.conv3, self.relu, self.pool, self.
conv4, self.relu)
self.bilinear = torch.nn.Bilinear(co_attention_dim, co_attention_dim, 1
)
self.conv = nn.Conv2d(1, 1, kernel_size=3, padding=1)
self.sigmoid = nn.Sigmoid()
def forward(self, query_embed, prev_embed, refer_embed, refer_label,
ori_size=None, use_float16=True):
"""
Args:
query_embed:[height,width,embed_dim]
prev_embed:[height,width,embed_dim]
ref_embed:[height,width,embed_dim]
ref_label:[height,width,obj_nums]
Return:
x:[h,w,obj_num,1]
"""
h, w, _ = query_embed.size()
obj_nums = refer_label.size(2)
query_embed = query_embed.permute(2, 0, 1)
prev_embed = prev_embed.permute(2, 1, 0)
refer_embed = refer_embed.permute(2, 1, 0)
refer_label = refer_label.permute(2, 0, 1)
r_attention = self.leak_relu(torch.bmm(refer_embed, query_embed)
).unsqueeze(0)
p_attention = self.leak_relu(torch.bmm(prev_embed, query_embed)
).unsqueeze(0)
r_attention = self.gate1(r_attention).squeeze(0)
p_attention = self.gate2(p_attention).squeeze(0)
attention_h = r_attention.size(1)
attention_w = r_attention.size(2)
r_attention = r_attention.reshape(attention_h * attention_w, -1)
p_attention = p_attention.reshape(attention_h * attention_w, -1)
attention = self.bilinear(r_attention, p_attention).reshape(-1, 1,
attention_h, attention_w)
if attention.size(2) != 64 or attention.size(3) != 64:
attention = F.interpolate(attention, size=[64, 64], mode=
'bilinear', align_corners=True)
attention = self.conv(attention)
attention = F.interpolate(attention, size=(h, w), mode='bilinear',
align_corners=True).squeeze(0)
x = torch.mul(attention, refer_label).unsqueeze(-1).permute(1, 2, 0, 3)
attention = attention.permute(1, 2, 0)
x = self.sigmoid(x)
if ori_size is not None:
x = x.view(h, w, obj_nums, 1).permute(2, 3, 0, 1)
x = F.interpolate(x, size=ori_size, mode='bilinear',
align_corners=True).permute(2, 3, 0, 1).view(ori_size[0],
ori_size[1], obj_nums, 1)
if use_float16:
x = x.float()
return x
def get_inputs():
return [torch.rand([4, 4, 4]), torch.rand([4, 4, 4]), torch.rand([4, 4,
4]), torch.rand([4, 4, 4])]
def get_init_inputs():
return [[], {'in_dim': 4, 'co_attention_dim': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_leaky_relu_0(in_out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_out_ptr0 + x0, xmask)
tmp1 = 0.0
tmp2 = tmp0 > tmp1
tmp3 = 0.01
tmp4 = tmp0 * tmp3
tmp5 = tl.where(tmp2, tmp0, tmp4)
tl.store(in_out_ptr0 + x0, tmp5, xmask)
@triton.jit
def triton_poi_fused_convolution_relu_1(in_out_ptr0, in_ptr0, xnumel,
XBLOCK: tl.constexpr):
xnumel = 1024
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 16
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x1, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tl.store(in_out_ptr0 + x2, tmp4, xmask)
@triton.jit
def triton_poi_fused_max_pool2d_with_indices_2(in_ptr0, out_ptr0, out_ptr1,
xnumel, XBLOCK: tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 2
x1 = xindex // 2
x2 = xindex
tmp0 = tl.load(in_ptr0 + (2 * x0 + 8 * x1), xmask, eviction_policy=
'evict_last')
tmp1 = tl.load(in_ptr0 + (1 + 2 * x0 + 8 * x1), xmask, eviction_policy=
'evict_last')
tmp3 = tl.load(in_ptr0 + (4 + 2 * x0 + 8 * x1), xmask, eviction_policy=
'evict_last')
tmp5 = tl.load(in_ptr0 + (5 + 2 * x0 + 8 * x1), xmask, eviction_policy=
'evict_last')
tmp2 = triton_helpers.maximum(tmp1, tmp0)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp6 = triton_helpers.maximum(tmp5, tmp4)
tmp7 = tmp1 > tmp0
tmp8 = tl.full([1], 1, tl.int8)
tmp9 = tl.full([1], 0, tl.int8)
tmp10 = tl.where(tmp7, tmp8, tmp9)
tmp11 = tmp3 > tmp2
tmp12 = tl.full([1], 2, tl.int8)
tmp13 = tl.where(tmp11, tmp12, tmp10)
tmp14 = tmp5 > tmp4
tmp15 = tl.full([1], 3, tl.int8)
tmp16 = tl.where(tmp14, tmp15, tmp13)
tl.store(out_ptr0 + x2, tmp6, xmask)
tl.store(out_ptr1 + x2, tmp16, xmask)
@triton.jit
def triton_poi_fused_convolution_relu_threshold_backward_3(in_out_ptr0,
in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x1, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(in_out_ptr0 + x2, tmp4, xmask)
tl.store(out_ptr0 + x2, tmp6, xmask)
@triton.jit
def triton_poi_fused__to_copy_4(out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = x0
tmp1 = tmp0.to(tl.float32)
tmp2 = 0.015873015873015872
tmp3 = tmp1 * tmp2
tmp4 = 0.0
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp6 = tmp5.to(tl.int32)
tl.store(out_ptr0 + x0, tmp6, xmask)
@triton.jit
def triton_poi_fused_add_clamp_5(out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = x0
tmp1 = tmp0.to(tl.float32)
tmp2 = 0.015873015873015872
tmp3 = tmp1 * tmp2
tmp4 = 0.0
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp6 = tmp5.to(tl.int32)
tmp7 = tl.full([1], 1, tl.int64)
tmp8 = tmp6 + tmp7
tmp9 = triton_helpers.minimum(tmp8, tmp7)
tl.store(out_ptr0 + x0, tmp9, xmask)
@triton.jit
def triton_poi_fused__to_copy_arange_clamp_mul_sub_6(out_ptr0, xnumel,
XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = x0
tmp1 = tmp0.to(tl.float32)
tmp2 = 0.015873015873015872
tmp3 = tmp1 * tmp2
tmp4 = 0.0
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp6 = tmp5.to(tl.int32)
tmp7 = tmp6.to(tl.float32)
tmp8 = tmp5 - tmp7
tmp9 = triton_helpers.maximum(tmp8, tmp4)
tmp10 = 1.0
tmp11 = triton_helpers.minimum(tmp9, tmp10)
tl.store(out_ptr0 + x0, tmp11, xmask)
@triton.jit
def triton_poi_fused__unsafe_index_add_mul_sub_7(in_out_ptr0, in_ptr0,
in_ptr1, in_ptr2, in_ptr3, in_ptr4, in_ptr5, in_ptr6, in_ptr7, xnumel,
XBLOCK: tl.constexpr):
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
tl.full([XBLOCK], True, tl.int1)
x1 = xindex // 64
x0 = xindex % 64
x2 = xindex
tmp0 = tl.load(in_ptr0 + x1, None, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr1 + x0, None, eviction_policy='evict_last')
tmp10 = tl.load(in_ptr3 + 0)
tmp11 = tl.broadcast_to(tmp10, [XBLOCK])
tmp13 = tl.load(in_ptr4 + x0, None, eviction_policy='evict_last')
tmp20 = tl.load(in_ptr5 + x0, None, eviction_policy='evict_last')
tmp23 = tl.load(in_ptr6 + x1, None, eviction_policy='evict_last')
tmp35 = tl.load(in_ptr7 + x1, None, eviction_policy='evict_last')
tmp1 = tl.full([XBLOCK], 2, tl.int32)
tmp2 = tmp0 + tmp1
tmp3 = tmp0 < 0
tmp4 = tl.where(tmp3, tmp2, tmp0)
tmp6 = tmp5 + tmp1
tmp7 = tmp5 < 0
tmp8 = tl.where(tmp7, tmp6, tmp5)
tmp9 = tl.load(in_ptr2 + (tmp8 + 2 * tmp4), None, eviction_policy=
'evict_last')
tmp12 = tmp9 + tmp11
tmp14 = tmp13 + tmp1
tmp15 = tmp13 < 0
tmp16 = tl.where(tmp15, tmp14, tmp13)
tmp17 = tl.load(in_ptr2 + (tmp16 + 2 * tmp4), None, eviction_policy=
'evict_last')
tmp18 = tmp17 + tmp11
tmp19 = tmp18 - tmp12
tmp21 = tmp19 * tmp20
tmp22 = tmp12 + tmp21
tmp24 = tmp23 + tmp1
tmp25 = tmp23 < 0
tmp26 = tl.where(tmp25, tmp24, tmp23)
tmp27 = tl.load(in_ptr2 + (tmp8 + 2 * tmp26), None, eviction_policy=
'evict_last')
tmp28 = tmp27 + tmp11
tmp29 = tl.load(in_ptr2 + (tmp16 + 2 * tmp26), None, eviction_policy=
'evict_last')
tmp30 = tmp29 + tmp11
tmp31 = tmp30 - tmp28
tmp32 = tmp31 * tmp20
tmp33 = tmp28 + tmp32
tmp34 = tmp33 - tmp22
tmp36 = tmp34 * tmp35
tmp37 = tmp22 + tmp36
tl.store(in_out_ptr0 + x2, tmp37, None)
@triton.jit
def triton_poi_fused__to_copy_8(out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 4
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = x0
tmp1 = tmp0.to(tl.float32)
tmp2 = 21.0
tmp3 = tmp1 * tmp2
tmp4 = 0.0
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp6 = tmp5.to(tl.int32)
tl.store(out_ptr0 + x0, tmp6, xmask)
@triton.jit
def triton_poi_fused_add_clamp_9(out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 4
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = x0
tmp1 = tmp0.to(tl.float32)
tmp2 = 21.0
tmp3 = tmp1 * tmp2
tmp4 = 0.0
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp6 = tmp5.to(tl.int32)
tmp7 = tl.full([1], 1, tl.int64)
tmp8 = tmp6 + tmp7
tmp9 = tl.full([1], 63, tl.int64)
tmp10 = triton_helpers.minimum(tmp8, tmp9)
tl.store(out_ptr0 + x0, tmp10, xmask)
@triton.jit
def triton_poi_fused__to_copy_arange_clamp_mul_sub_10(out_ptr0, xnumel,
XBLOCK: tl.constexpr):
xnumel = 4
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = x0
tmp1 = tmp0.to(tl.float32)
tmp2 = 21.0
tmp3 = tmp1 * tmp2
tmp4 = 0.0
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp6 = tmp5.to(tl.int32)
tmp7 = tmp6.to(tl.float32)
tmp8 = tmp5 - tmp7
tmp9 = triton_helpers.maximum(tmp8, tmp4)
tmp10 = 1.0
tmp11 = triton_helpers.minimum(tmp9, tmp10)
tl.store(out_ptr0 + x0, tmp11, xmask)
@triton.jit
def triton_poi_fused__unsafe_index_add_convolution_mul_sub_11(in_ptr0,
in_ptr1, in_ptr2, in_ptr3, in_ptr4, in_ptr5, in_ptr6, in_ptr7, out_ptr0,
out_ptr1, xnumel, XBLOCK: tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = xindex // 4
x0 = xindex % 4
x2 = xindex
tmp0 = tl.load(in_ptr0 + x1, xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr1 + x0, xmask, eviction_policy='evict_last')
tmp10 = tl.load(in_ptr3 + 0)
tmp11 = tl.broadcast_to(tmp10, [XBLOCK])
tmp13 = tl.load(in_ptr4 + x0, xmask, eviction_policy='evict_last')
tmp20 = tl.load(in_ptr5 + x0, xmask, eviction_policy='evict_last')
tmp23 = tl.load(in_ptr6 + x1, xmask, eviction_policy='evict_last')
tmp35 = tl.load(in_ptr7 + x1, xmask, eviction_policy='evict_last')
tmp1 = tl.full([XBLOCK], 64, tl.int32)
tmp2 = tmp0 + tmp1
tmp3 = tmp0 < 0
tmp4 = tl.where(tmp3, tmp2, tmp0)
tmp6 = tmp5 + tmp1
tmp7 = tmp5 < 0
tmp8 = tl.where(tmp7, tmp6, tmp5)
tmp9 = tl.load(in_ptr2 + (tmp8 + 64 * tmp4), xmask, eviction_policy=
'evict_last')
tmp12 = tmp9 + tmp11
tmp14 = tmp13 + tmp1
tmp15 = tmp13 < 0
tmp16 = tl.where(tmp15, tmp14, tmp13)
tmp17 = tl.load(in_ptr2 + (tmp16 + 64 * tmp4), xmask, eviction_policy=
'evict_last')
tmp18 = tmp17 + tmp11
tmp19 = tmp18 - tmp12
tmp21 = tmp19 * tmp20
tmp22 = tmp12 + tmp21
tmp24 = tmp23 + tmp1
tmp25 = tmp23 < 0
tmp26 = tl.where(tmp25, tmp24, tmp23)
tmp27 = tl.load(in_ptr2 + (tmp8 + 64 * tmp26), xmask, eviction_policy=
'evict_last')
tmp28 = tmp27 + tmp11
tmp29 = tl.load(in_ptr2 + (tmp16 + 64 * tmp26), xmask, eviction_policy=
'evict_last')
tmp30 = tmp29 + tmp11
tmp31 = tmp30 - tmp28
tmp32 = tmp31 * tmp20
tmp33 = tmp28 + tmp32
tmp34 = tmp33 - tmp22
tmp36 = tmp34 * tmp35
tl.store(out_ptr0 + x2, tmp22, xmask)
tl.store(out_ptr1 + x2, tmp36, xmask)
@triton.jit
def triton_poi_fused_sigmoid_12(in_ptr0, in_ptr1, in_ptr2, out_ptr0, xnumel,
XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = xindex // 4
x2 = xindex
tmp0 = tl.load(in_ptr0 + x1, xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + x1, xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr2 + x2, xmask)
tmp2 = tmp0 + tmp1
tmp4 = tmp2 * tmp3
tmp5 = tl.sigmoid(tmp4)
tl.store(out_ptr0 + x2, tmp5, xmask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7, primals_8, primals_9, primals_10, primals_11, primals_12,
primals_13, primals_14, primals_15, primals_16) = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_2, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_3, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_4, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_5, (64, 4, 3, 3), (36, 9, 3, 1))
assert_size_stride(primals_6, (64,), (1,))
assert_size_stride(primals_7, (4, 64, 3, 3), (576, 9, 3, 1))
assert_size_stride(primals_8, (4,), (1,))
assert_size_stride(primals_9, (64, 4, 3, 3), (36, 9, 3, 1))
assert_size_stride(primals_10, (64,), (1,))
assert_size_stride(primals_11, (4, 64, 3, 3), (576, 9, 3, 1))
assert_size_stride(primals_12, (4,), (1,))
assert_size_stride(primals_13, (1, 4, 4), (16, 4, 1))
assert_size_stride(primals_14, (1,), (1,))
assert_size_stride(primals_15, (1, 1, 3, 3), (9, 9, 3, 1))
assert_size_stride(primals_16, (1,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
extern_kernels.bmm(reinterpret_tensor(primals_4, (4, 4, 4), (1, 4,
16), 0), reinterpret_tensor(primals_1, (4, 4, 4), (1, 16, 4), 0
), out=buf0)
del primals_4
buf1 = buf0
del buf0
get_raw_stream(0)
triton_poi_fused_leaky_relu_0[grid(64)](buf1, 64, XBLOCK=64,
num_warps=1, num_stages=1)
buf2 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
extern_kernels.bmm(reinterpret_tensor(primals_3, (4, 4, 4), (1, 4,
16), 0), reinterpret_tensor(primals_1, (4, 4, 4), (1, 16, 4), 0
), out=buf2)
del primals_1
del primals_3
buf3 = buf2
del buf2
triton_poi_fused_leaky_relu_0[grid(64)](buf3, 64, XBLOCK=64,
num_warps=1, num_stages=1)
buf4 = extern_kernels.convolution(reinterpret_tensor(buf1, (1, 4, 4,
4), (0, 16, 4, 1), 0), primals_5, stride=(1, 1), padding=(1, 1),
dilation=(1, 1), transposed=False, output_padding=(0, 0),
groups=1, bias=None)
assert_size_stride(buf4, (1, 64, 4, 4), (1024, 16, 4, 1))
buf5 = buf4
del buf4
triton_poi_fused_convolution_relu_1[grid(1024)](buf5, primals_6,
1024, XBLOCK=256, num_warps=4, num_stages=1)
del primals_6
buf6 = empty_strided_cuda((1, 64, 2, 2), (256, 4, 2, 1), torch.float32)
buf7 = empty_strided_cuda((1, 64, 2, 2), (256, 4, 2, 1), torch.int8)
triton_poi_fused_max_pool2d_with_indices_2[grid(256)](buf5, buf6,
buf7, 256, XBLOCK=128, num_warps=4, num_stages=1)
buf8 = extern_kernels.convolution(buf6, primals_7, stride=(1, 1),
padding=(1, 1), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf8, (1, 4, 2, 2), (16, 4, 2, 1))
buf9 = extern_kernels.convolution(reinterpret_tensor(buf3, (1, 4, 4,
4), (0, 16, 4, 1), 0), primals_9, stride=(1, 1), padding=(1, 1),
dilation=(1, 1), transposed=False, output_padding=(0, 0),
groups=1, bias=None)
assert_size_stride(buf9, (1, 64, 4, 4), (1024, 16, 4, 1))
buf10 = buf9
del buf9
triton_poi_fused_convolution_relu_1[grid(1024)](buf10, primals_10,
1024, XBLOCK=256, num_warps=4, num_stages=1)
del primals_10
buf11 = empty_strided_cuda((1, 64, 2, 2), (256, 4, 2, 1), torch.float32
)
buf12 = empty_strided_cuda((1, 64, 2, 2), (256, 4, 2, 1), torch.int8)
triton_poi_fused_max_pool2d_with_indices_2[grid(256)](buf10, buf11,
buf12, 256, XBLOCK=128, num_warps=4, num_stages=1)
buf13 = extern_kernels.convolution(buf11, primals_11, stride=(1, 1),
padding=(1, 1), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf13, (1, 4, 2, 2), (16, 4, 2, 1))
buf14 = buf8
del buf8
buf38 = empty_strided_cuda((1, 4, 2, 2), (16, 4, 2, 1), torch.bool)
triton_poi_fused_convolution_relu_threshold_backward_3[grid(16)](buf14,
primals_8, buf38, 16, XBLOCK=16, num_warps=1, num_stages=1)
del primals_8
buf15 = buf13
del buf13
buf37 = empty_strided_cuda((1, 4, 2, 2), (16, 4, 2, 1), torch.bool)
triton_poi_fused_convolution_relu_threshold_backward_3[grid(16)](buf15,
primals_12, buf37, 16, XBLOCK=16, num_warps=1, num_stages=1)
del primals_12
buf16 = torch.ops.aten._trilinear.default(reinterpret_tensor(buf14,
(4, 4), (4, 1), 0), primals_13, reinterpret_tensor(buf15, (4, 4
), (4, 1), 0), [1, 3], [0], [1, 2], [2, 3])
buf17 = buf16
del buf16
buf18 = empty_strided_cuda((64, 1), (1, 1), torch.int64)
triton_poi_fused__to_copy_4[grid(64)](buf18, 64, XBLOCK=64,
num_warps=1, num_stages=1)
buf19 = empty_strided_cuda((64, 1), (1, 1), torch.int64)
triton_poi_fused_add_clamp_5[grid(64)](buf19, 64, XBLOCK=64,
num_warps=1, num_stages=1)
buf20 = empty_strided_cuda((64,), (1,), torch.int64)
triton_poi_fused__to_copy_4[grid(64)](buf20, 64, XBLOCK=64,
num_warps=1, num_stages=1)
buf21 = empty_strided_cuda((64,), (1,), torch.int64)
triton_poi_fused_add_clamp_5[grid(64)](buf21, 64, XBLOCK=64,
num_warps=1, num_stages=1)
buf22 = empty_strided_cuda((64,), (1,), torch.float32)
triton_poi_fused__to_copy_arange_clamp_mul_sub_6[grid(64)](buf22,
64, XBLOCK=64, num_warps=1, num_stages=1)
buf24 = empty_strided_cuda((64, 1), (1, 1), torch.float32)
triton_poi_fused__to_copy_arange_clamp_mul_sub_6[grid(64)](buf24,
64, XBLOCK=64, num_warps=1, num_stages=1)
buf25 = empty_strided_cuda((1, 1, 64, 64), (4096, 4096, 64, 1),
torch.float32)
buf26 = buf25
del buf25
triton_poi_fused__unsafe_index_add_mul_sub_7[grid(4096)](buf26,
buf18, buf20, buf17, primals_14, buf21, buf22, buf19, buf24,
4096, XBLOCK=256, num_warps=4, num_stages=1)
del primals_14
buf27 = extern_kernels.convolution(buf26, primals_15, stride=(1, 1),
padding=(1, 1), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf27, (1, 1, 64, 64), (4096, 4096, 64, 1))
buf28 = empty_strided_cuda((4, 1), (1, 1), torch.int64)
triton_poi_fused__to_copy_8[grid(4)](buf28, 4, XBLOCK=4, num_warps=
1, num_stages=1)
buf29 = empty_strided_cuda((4, 1), (1, 1), torch.int64)
triton_poi_fused_add_clamp_9[grid(4)](buf29, 4, XBLOCK=4, num_warps
=1, num_stages=1)
buf30 = empty_strided_cuda((4,), (1,), torch.int64)
triton_poi_fused__to_copy_8[grid(4)](buf30, 4, XBLOCK=4, num_warps=
1, num_stages=1)
buf31 = empty_strided_cuda((4,), (1,), torch.int64)
triton_poi_fused_add_clamp_9[grid(4)](buf31, 4, XBLOCK=4, num_warps
=1, num_stages=1)
buf32 = reinterpret_tensor(buf17, (4,), (1,), 0)
del buf17
triton_poi_fused__to_copy_arange_clamp_mul_sub_10[grid(4)](buf32, 4,
XBLOCK=4, num_warps=1, num_stages=1)
buf34 = empty_strided_cuda((4, 1), (1, 1), torch.float32)
triton_poi_fused__to_copy_arange_clamp_mul_sub_10[grid(4)](buf34, 4,
XBLOCK=4, num_warps=1, num_stages=1)
buf33 = empty_strided_cuda((1, 1, 4, 4), (16, 16, 4, 1), torch.float32)
buf35 = empty_strided_cuda((1, 1, 4, 4), (16, 16, 4, 1), torch.float32)
triton_poi_fused__unsafe_index_add_convolution_mul_sub_11[grid(16)](
buf28, buf30, buf27, primals_16, buf31, buf32, buf29, buf34,
buf33, buf35, 16, XBLOCK=16, num_warps=1, num_stages=1)
del buf27
del primals_16
buf36 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 1), torch.float32)
triton_poi_fused_sigmoid_12[grid(64)](buf33, buf35, primals_2,
buf36, 64, XBLOCK=64, num_warps=1, num_stages=1)
del buf33
del buf35
return (buf36, primals_2, primals_5, primals_7, primals_9, primals_11,
primals_13, primals_15, reinterpret_tensor(buf1, (1, 4, 4, 4), (64,
16, 4, 1), 0), reinterpret_tensor(buf3, (1, 4, 4, 4), (64, 16, 4, 1
), 0), buf5, buf6, buf7, buf10, buf11, buf12, reinterpret_tensor(
buf14, (4, 4), (4, 1), 0), reinterpret_tensor(buf15, (4, 4), (4, 1),
0), buf18, buf19, buf20, buf21, buf22, buf24, buf26, buf28, buf29,
buf30, buf31, buf32, buf34, buf36, buf37, buf38)
class CO_AttentionNew(nn.Module):
def __init__(self, in_dim, co_attention_dim):
super(CO_AttentionNew, self).__init__()
self.leak_relu = nn.LeakyReLU()
self.relu = nn.ReLU()
self.conv1 = nn.Conv2d(in_dim, 64, kernel_size=3, padding=1)
self.conv2 = nn.Conv2d(64, co_attention_dim, kernel_size=3, padding=1)
self.pool = torch.nn.MaxPool2d(kernel_size=2)
self.gate1 = nn.Sequential(self.conv1, self.relu, self.pool, self.
conv2, self.relu)
self.conv3 = nn.Conv2d(in_dim, 64, kernel_size=3, padding=1)
self.conv4 = nn.Conv2d(64, co_attention_dim, kernel_size=3, padding=1)
self.gate2 = nn.Sequential(self.conv3, self.relu, self.pool, self.
conv4, self.relu)
self.bilinear = torch.nn.Bilinear(co_attention_dim, co_attention_dim, 1
)
self.conv = nn.Conv2d(1, 1, kernel_size=3, padding=1)
self.sigmoid = nn.Sigmoid()
def forward(self, input_0, input_1, input_2, input_3):
primals_5 = self.conv1.weight
primals_6 = self.conv1.bias
primals_7 = self.conv2.weight
primals_8 = self.conv2.bias
primals_9 = self.conv3.weight
primals_10 = self.conv3.bias
primals_11 = self.conv4.weight
primals_12 = self.conv4.bias
primals_13 = self.bilinear.weight
primals_14 = self.bilinear.bias
primals_15 = self.conv.weight
primals_16 = self.conv.bias
primals_1 = input_0
primals_2 = input_1
primals_3 = input_2
primals_4 = input_3
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7, primals_8, primals_9,
primals_10, primals_11, primals_12, primals_13, primals_14,
primals_15, primals_16])
return output[0]
| huanglf714/COMatchNet | CO_Attention | false | 6,858 | [
"Apache-2.0"
] | 1 | 79023f5be65d354eb9bdac026d7e0d73110bc4aa | https://github.com/huanglf714/COMatchNet/tree/79023f5be65d354eb9bdac026d7e0d73110bc4aa | import torch
import torch.nn as nn
import torch.nn.functional as F
class Model(nn.Module):
def __init__(self, in_dim, co_attention_dim):
super().__init__()
self.leak_relu = nn.LeakyReLU()
self.relu = nn.ReLU()
self.conv1 = nn.Conv2d(in_dim, 64, kernel_size=3, padding=1)
self.conv2 = nn.Conv2d(64, co_attention_dim, kernel_size=3, padding=1)
self.pool = torch.nn.MaxPool2d(kernel_size=2)
self.gate1 = nn.Sequential(self.conv1, self.relu, self.pool, self.
conv2, self.relu)
self.conv3 = nn.Conv2d(in_dim, 64, kernel_size=3, padding=1)
self.conv4 = nn.Conv2d(64, co_attention_dim, kernel_size=3, padding=1)
self.gate2 = nn.Sequential(self.conv3, self.relu, self.pool, self.
conv4, self.relu)
self.bilinear = torch.nn.Bilinear(co_attention_dim, co_attention_dim, 1
)
self.conv = nn.Conv2d(1, 1, kernel_size=3, padding=1)
self.sigmoid = nn.Sigmoid()
def forward(self, query_embed, prev_embed, refer_embed, refer_label,
ori_size=None, use_float16=True):
"""
Args:
query_embed:[height,width,embed_dim]
prev_embed:[height,width,embed_dim]
ref_embed:[height,width,embed_dim]
ref_label:[height,width,obj_nums]
Return:
x:[h,w,obj_num,1]
"""
h, w, _ = query_embed.size()
obj_nums = refer_label.size(2)
query_embed = query_embed.permute(2, 0, 1)
prev_embed = prev_embed.permute(2, 1, 0)
refer_embed = refer_embed.permute(2, 1, 0)
refer_label = refer_label.permute(2, 0, 1)
r_attention = self.leak_relu(torch.bmm(refer_embed, query_embed)
).unsqueeze(0)
p_attention = self.leak_relu(torch.bmm(prev_embed, query_embed)
).unsqueeze(0)
r_attention = self.gate1(r_attention).squeeze(0)
p_attention = self.gate2(p_attention).squeeze(0)
attention_h = r_attention.size(1)
attention_w = r_attention.size(2)
r_attention = r_attention.reshape(attention_h * attention_w, -1)
p_attention = p_attention.reshape(attention_h * attention_w, -1)
attention = self.bilinear(r_attention, p_attention).reshape(-1, 1,
attention_h, attention_w)
if attention.size(2) != 64 or attention.size(3) != 64:
attention = F.interpolate(attention, size=[64, 64], mode=
'bilinear', align_corners=True)
attention = self.conv(attention)
attention = F.interpolate(attention, size=(h, w), mode='bilinear',
align_corners=True).squeeze(0)
x = torch.mul(attention, refer_label).unsqueeze(-1).permute(1, 2, 0, 3)
attention = attention.permute(1, 2, 0)
x = self.sigmoid(x)
if ori_size is not None:
x = x.view(h, w, obj_nums, 1).permute(2, 3, 0, 1)
x = F.interpolate(x, size=ori_size, mode='bilinear',
align_corners=True).permute(2, 3, 0, 1).view(ori_size[0],
ori_size[1], obj_nums, 1)
if use_float16:
x = x.float()
return x
def get_inputs():
return [torch.rand([4, 4, 4]), torch.rand([4, 4, 4]), torch.rand([4, 4,
4]), torch.rand([4, 4, 4])]
def get_init_inputs():
return [4, 4]
|
ActorDiscrete | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/v7/cv7zazascu4rpkkwoxbiwk6c2le2e6wshdhae73bmaoapelvwguv.py
# Topologically Sorted Source Nodes: [a], Original ATen: [aten.relu, aten.threshold_backward]
# Source node to ATen node mapping:
# a => relu
# Graph fragment:
# %relu : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%view_1,), kwargs = {})
# %le_1 : [num_users=1] = call_function[target=torch.ops.aten.le.Scalar](args = (%relu, 0), kwargs = {})
triton_poi_fused_relu_threshold_backward_0 = async_compile.triton('triton_poi_fused_relu_threshold_backward_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[1024],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*i1', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_relu_threshold_backward_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_relu_threshold_backward_0(in_out_ptr0, in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 1024
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 16
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(in_out_ptr0 + (x2), tmp4, xmask)
tl.store(out_ptr0 + (x2), tmp6, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/xk/cxkugsynlmnyrjhah42fewrhwovuvurnuv2qimo2qhxq27wjmq7q.py
# Topologically Sorted Source Nodes: [a_2], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# a_2 => amax, exp, sub
# Graph fragment:
# %amax : [num_users=1] = call_function[target=torch.ops.aten.amax.default](args = (%view_5, [1], True), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%view_5, %amax), kwargs = {})
# %exp : [num_users=2] = call_function[target=torch.ops.aten.exp.default](args = (%sub,), kwargs = {})
triton_poi_fused__softmax_1 = async_compile.triton('triton_poi_fused__softmax_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_1(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x0 = xindex % 16
x2 = (xindex // 64)
tmp0 = tl.load(in_ptr0 + (x3), xmask)
tmp1 = tl.load(in_ptr0 + (x0 + (64*x2)), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (16 + x0 + (64*x2)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (32 + x0 + (64*x2)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (48 + x0 + (64*x2)), xmask, eviction_policy='evict_last')
tmp3 = triton_helpers.maximum(tmp1, tmp2)
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp7 = triton_helpers.maximum(tmp5, tmp6)
tmp8 = tmp0 - tmp7
tmp9 = tl_math.exp(tmp8)
tl.store(out_ptr0 + (x3), tmp9, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/jf/cjfzp64ny4hf7wdw5wptah3hqv5fcsh5rrw4brz7uxcy6ad57n7h.py
# Topologically Sorted Source Nodes: [a_2], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# a_2 => div, sum_1
# Graph fragment:
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%exp, [1], True), kwargs = {})
# %div : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%exp, %sum_1), kwargs = {})
triton_poi_fused__softmax_2 = async_compile.triton('triton_poi_fused__softmax_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_2(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x0 = xindex % 16
x2 = (xindex // 64)
tmp0 = tl.load(in_ptr0 + (x3), xmask)
tmp1 = tl.load(in_ptr0 + (x0 + (64*x2)), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (16 + x0 + (64*x2)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (32 + x0 + (64*x2)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (48 + x0 + (64*x2)), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + (x3), tmp8, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7 = args
args.clear()
assert_size_stride(primals_1, (16, 4), (4, 1))
assert_size_stride(primals_2, (16, ), (1, ))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_4, (16, 16), (16, 1))
assert_size_stride(primals_5, (16, ), (1, ))
assert_size_stride(primals_6, (4, 16), (16, 1))
assert_size_stride(primals_7, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 16), (16, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(primals_3, (64, 4), (4, 1), 0), reinterpret_tensor(primals_1, (4, 16), (1, 4), 0), out=buf0)
del primals_1
buf1 = reinterpret_tensor(buf0, (4, 4, 4, 16), (256, 64, 16, 1), 0); del buf0 # reuse
buf8 = empty_strided_cuda((4, 4, 4, 16), (256, 64, 16, 1), torch.bool)
# Topologically Sorted Source Nodes: [a], Original ATen: [aten.relu, aten.threshold_backward]
stream0 = get_raw_stream(0)
triton_poi_fused_relu_threshold_backward_0.run(buf1, primals_2, buf8, 1024, grid=grid(1024), stream=stream0)
del primals_2
buf2 = empty_strided_cuda((64, 16), (16, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(buf1, (64, 16), (16, 1), 0), reinterpret_tensor(primals_4, (16, 16), (1, 16), 0), out=buf2)
buf3 = reinterpret_tensor(buf2, (4, 4, 4, 16), (256, 64, 16, 1), 0); del buf2 # reuse
buf7 = empty_strided_cuda((4, 4, 4, 16), (256, 64, 16, 1), torch.bool)
# Topologically Sorted Source Nodes: [a_1], Original ATen: [aten.relu, aten.threshold_backward]
triton_poi_fused_relu_threshold_backward_0.run(buf3, primals_5, buf7, 1024, grid=grid(1024), stream=stream0)
del primals_5
buf4 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [linear_2], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_7, reinterpret_tensor(buf3, (64, 16), (16, 1), 0), reinterpret_tensor(primals_6, (16, 4), (1, 16), 0), alpha=1, beta=1, out=buf4)
del primals_7
buf5 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [a_2], Original ATen: [aten._softmax]
triton_poi_fused__softmax_1.run(buf4, buf5, 256, grid=grid(256), stream=stream0)
buf6 = reinterpret_tensor(buf4, (4, 4, 4, 4), (64, 16, 4, 1), 0); del buf4 # reuse
# Topologically Sorted Source Nodes: [a_2], Original ATen: [aten._softmax]
triton_poi_fused__softmax_2.run(buf5, buf6, 256, grid=grid(256), stream=stream0)
del buf5
return (buf6, reinterpret_tensor(primals_3, (64, 4), (4, 1), 0), reinterpret_tensor(buf1, (64, 16), (16, 1), 0), reinterpret_tensor(buf3, (64, 16), (16, 1), 0), buf6, primals_6, buf7, primals_4, buf8, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((16, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((16, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((16, 16), (16, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((16, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((4, 16), (16, 1), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch as t
import torch.nn as nn
class ActorDiscrete(nn.Module):
def __init__(self, state_dim, action_dim):
super().__init__()
self.fc1 = nn.Linear(state_dim, 16)
self.fc2 = nn.Linear(16, 16)
self.fc3 = nn.Linear(16, action_dim)
def forward(self, state):
a = t.relu(self.fc1(state))
a = t.relu(self.fc2(a))
a = t.softmax(self.fc3(a), dim=1)
return a
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'state_dim': 4, 'action_dim': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import math as tl_math
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_relu_threshold_backward_0(in_out_ptr0, in_ptr0,
out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 1024
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 16
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(in_out_ptr0 + x2, tmp4, xmask)
tl.store(out_ptr0 + x2, tmp6, xmask)
@triton.jit
def triton_poi_fused__softmax_1(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x0 = xindex % 16
x2 = xindex // 64
tmp0 = tl.load(in_ptr0 + x3, xmask)
tmp1 = tl.load(in_ptr0 + (x0 + 64 * x2), xmask, eviction_policy=
'evict_last')
tmp2 = tl.load(in_ptr0 + (16 + x0 + 64 * x2), xmask, eviction_policy=
'evict_last')
tmp4 = tl.load(in_ptr0 + (32 + x0 + 64 * x2), xmask, eviction_policy=
'evict_last')
tmp6 = tl.load(in_ptr0 + (48 + x0 + 64 * x2), xmask, eviction_policy=
'evict_last')
tmp3 = triton_helpers.maximum(tmp1, tmp2)
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp7 = triton_helpers.maximum(tmp5, tmp6)
tmp8 = tmp0 - tmp7
tmp9 = tl_math.exp(tmp8)
tl.store(out_ptr0 + x3, tmp9, xmask)
@triton.jit
def triton_poi_fused__softmax_2(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x0 = xindex % 16
x2 = xindex // 64
tmp0 = tl.load(in_ptr0 + x3, xmask)
tmp1 = tl.load(in_ptr0 + (x0 + 64 * x2), xmask, eviction_policy=
'evict_last')
tmp2 = tl.load(in_ptr0 + (16 + x0 + 64 * x2), xmask, eviction_policy=
'evict_last')
tmp4 = tl.load(in_ptr0 + (32 + x0 + 64 * x2), xmask, eviction_policy=
'evict_last')
tmp6 = tl.load(in_ptr0 + (48 + x0 + 64 * x2), xmask, eviction_policy=
'evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + x3, tmp8, xmask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7) = args
args.clear()
assert_size_stride(primals_1, (16, 4), (4, 1))
assert_size_stride(primals_2, (16,), (1,))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_4, (16, 16), (16, 1))
assert_size_stride(primals_5, (16,), (1,))
assert_size_stride(primals_6, (4, 16), (16, 1))
assert_size_stride(primals_7, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 16), (16, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_3, (64, 4), (4, 1), 0),
reinterpret_tensor(primals_1, (4, 16), (1, 4), 0), out=buf0)
del primals_1
buf1 = reinterpret_tensor(buf0, (4, 4, 4, 16), (256, 64, 16, 1), 0)
del buf0
buf8 = empty_strided_cuda((4, 4, 4, 16), (256, 64, 16, 1), torch.bool)
get_raw_stream(0)
triton_poi_fused_relu_threshold_backward_0[grid(1024)](buf1,
primals_2, buf8, 1024, XBLOCK=128, num_warps=4, num_stages=1)
del primals_2
buf2 = empty_strided_cuda((64, 16), (16, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(buf1, (64, 16), (16, 1), 0),
reinterpret_tensor(primals_4, (16, 16), (1, 16), 0), out=buf2)
buf3 = reinterpret_tensor(buf2, (4, 4, 4, 16), (256, 64, 16, 1), 0)
del buf2
buf7 = empty_strided_cuda((4, 4, 4, 16), (256, 64, 16, 1), torch.bool)
triton_poi_fused_relu_threshold_backward_0[grid(1024)](buf3,
primals_5, buf7, 1024, XBLOCK=128, num_warps=4, num_stages=1)
del primals_5
buf4 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
extern_kernels.addmm(primals_7, reinterpret_tensor(buf3, (64, 16),
(16, 1), 0), reinterpret_tensor(primals_6, (16, 4), (1, 16), 0),
alpha=1, beta=1, out=buf4)
del primals_7
buf5 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
triton_poi_fused__softmax_1[grid(256)](buf4, buf5, 256, XBLOCK=128,
num_warps=4, num_stages=1)
buf6 = reinterpret_tensor(buf4, (4, 4, 4, 4), (64, 16, 4, 1), 0)
del buf4
triton_poi_fused__softmax_2[grid(256)](buf5, buf6, 256, XBLOCK=256,
num_warps=4, num_stages=1)
del buf5
return buf6, reinterpret_tensor(primals_3, (64, 4), (4, 1), 0
), reinterpret_tensor(buf1, (64, 16), (16, 1), 0), reinterpret_tensor(
buf3, (64, 16), (16, 1), 0), buf6, primals_6, buf7, primals_4, buf8
class ActorDiscreteNew(nn.Module):
def __init__(self, state_dim, action_dim):
super().__init__()
self.fc1 = nn.Linear(state_dim, 16)
self.fc2 = nn.Linear(16, 16)
self.fc3 = nn.Linear(16, action_dim)
def forward(self, input_0):
primals_1 = self.fc1.weight
primals_2 = self.fc1.bias
primals_4 = self.fc2.weight
primals_5 = self.fc2.bias
primals_6 = self.fc3.weight
primals_7 = self.fc3.bias
primals_3 = input_0
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7])
return output[0]
| ikamensh/machin | ActorDiscrete | false | 6,859 | [
"MIT"
] | 1 | af7b423c47bc1412530cf6c96c11bd3af9b3e239 | https://github.com/ikamensh/machin/tree/af7b423c47bc1412530cf6c96c11bd3af9b3e239 | import torch
import torch as t
import torch.nn as nn
class Model(nn.Module):
def __init__(self, state_dim, action_dim):
super().__init__()
self.fc1 = nn.Linear(state_dim, 16)
self.fc2 = nn.Linear(16, 16)
self.fc3 = nn.Linear(16, action_dim)
def forward(self, state):
a = t.relu(self.fc1(state))
a = t.relu(self.fc2(a))
a = t.softmax(self.fc3(a), dim=1)
return a
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4, 4]
|
A2CActorDisc | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/wn/cwnbny2g2c4hzf4klds4ngvg6nqtj6s4fbxcgkvzmot3mq2igc6c.py
# Topologically Sorted Source Nodes: [a], Original ATen: [aten.relu]
# Source node to ATen node mapping:
# a => relu
# Graph fragment:
# %add_tensor_1 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mm_default_1, %primals_2), kwargs = {})
# %relu : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%add_tensor_1,), kwargs = {})
triton_poi_fused_relu_0 = async_compile.triton('triton_poi_fused_relu_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_relu_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_relu_0(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 16
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tl.store(in_out_ptr0 + (x2), tmp4, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/ts/ctscnzvbagjv4t25zui245b3recij5udu7nvujnr5rixcyo7elc6.py
# Topologically Sorted Source Nodes: [probs], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# probs => amax, exp, sub
# Graph fragment:
# %amax : [num_users=1] = call_function[target=torch.ops.aten.amax.default](args = (%addmm_2, [1], True), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%addmm_2, %amax), kwargs = {})
# %exp : [num_users=2] = call_function[target=torch.ops.aten.exp.default](args = (%sub,), kwargs = {})
triton_poi_fused__softmax_1 = async_compile.triton('triton_poi_fused__softmax_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_1(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp3 = triton_helpers.maximum(tmp1, tmp2)
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp7 = triton_helpers.maximum(tmp5, tmp6)
tmp8 = tmp0 - tmp7
tmp9 = tl_math.exp(tmp8)
tl.store(out_ptr0 + (x2), tmp9, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/k6/ck6fz3qsfeqgn5jtm4ugikmu7cwvvlq3jpttijbb5kdniicwtyz6.py
# Topologically Sorted Source Nodes: [probs], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# probs => div, sum_1
# Graph fragment:
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%exp, [1], True), kwargs = {})
# %div : [num_users=2] = call_function[target=torch.ops.aten.div.Tensor](args = (%exp, %sum_1), kwargs = {})
triton_poi_fused__softmax_2 = async_compile.triton('triton_poi_fused__softmax_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_2(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + (x2), tmp8, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7 = args
args.clear()
assert_size_stride(primals_1, (16, 4), (4, 1))
assert_size_stride(primals_2, (16, ), (1, ))
assert_size_stride(primals_3, (4, 4), (4, 1))
assert_size_stride(primals_4, (16, 16), (16, 1))
assert_size_stride(primals_5, (16, ), (1, ))
assert_size_stride(primals_6, (4, 16), (16, 1))
assert_size_stride(primals_7, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 16), (16, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(primals_3, reinterpret_tensor(primals_1, (4, 16), (1, 4), 0), out=buf0)
del primals_1
buf1 = buf0; del buf0 # reuse
# Topologically Sorted Source Nodes: [a], Original ATen: [aten.relu]
stream0 = get_raw_stream(0)
triton_poi_fused_relu_0.run(buf1, primals_2, 64, grid=grid(64), stream=stream0)
del primals_2
buf2 = empty_strided_cuda((4, 16), (16, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(buf1, reinterpret_tensor(primals_4, (16, 16), (1, 16), 0), out=buf2)
buf3 = buf2; del buf2 # reuse
# Topologically Sorted Source Nodes: [a_1], Original ATen: [aten.relu]
triton_poi_fused_relu_0.run(buf3, primals_5, 64, grid=grid(64), stream=stream0)
del primals_5
buf4 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [linear_2], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_7, buf3, reinterpret_tensor(primals_6, (16, 4), (1, 16), 0), alpha=1, beta=1, out=buf4)
del primals_7
buf5 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [probs], Original ATen: [aten._softmax]
triton_poi_fused__softmax_1.run(buf4, buf5, 16, grid=grid(16), stream=stream0)
buf6 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [probs], Original ATen: [aten._softmax]
triton_poi_fused__softmax_2.run(buf5, buf6, 16, grid=grid(16), stream=stream0)
buf7 = buf5; del buf5 # reuse
# Topologically Sorted Source Nodes: [sum_1, truediv], Original ATen: [aten.sum, aten.div]
triton_poi_fused__softmax_2.run(buf6, buf7, 16, grid=grid(16), stream=stream0)
del buf6
# Topologically Sorted Source Nodes: [multinomial], Original ATen: [aten.multinomial]
buf8 = torch.ops.aten.multinomial.default(buf7, 1, True)
buf9 = buf8
del buf8
return (reinterpret_tensor(buf9, (4, ), (1, ), 0), buf7, primals_3, buf1, buf3, buf4, primals_6, primals_4, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((16, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((16, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((16, 16), (16, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((16, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((4, 16), (16, 1), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
from torch.distributions import Categorical
import torch as t
import torch.nn as nn
class A2CActorDisc(nn.Module):
def __init__(self, state_dim, action_num):
super().__init__()
self.fc1 = nn.Linear(state_dim, 16)
self.fc2 = nn.Linear(16, 16)
self.fc3 = nn.Linear(16, action_num)
def forward(self, state, action=None):
a = t.relu(self.fc1(state))
a = t.relu(self.fc2(a))
probs = t.softmax(self.fc3(a), dim=1)
dist = Categorical(probs=probs)
act = action if action is not None else dist.sample()
act_entropy = dist.entropy()
act_log_prob = dist.log_prob(act.flatten())
return act, act_log_prob, act_entropy
def get_inputs():
return [torch.rand([4, 4])]
def get_init_inputs():
return [[], {'state_dim': 4, 'action_num': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import math as tl_math
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_relu_0(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 16
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tl.store(in_out_ptr0 + x2, tmp4, xmask)
@triton.jit
def triton_poi_fused__softmax_1(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp3 = triton_helpers.maximum(tmp1, tmp2)
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp7 = triton_helpers.maximum(tmp5, tmp6)
tmp8 = tmp0 - tmp7
tmp9 = tl_math.exp(tmp8)
tl.store(out_ptr0 + x2, tmp9, xmask)
@triton.jit
def triton_poi_fused__softmax_2(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + x2, tmp8, xmask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7) = args
args.clear()
assert_size_stride(primals_1, (16, 4), (4, 1))
assert_size_stride(primals_2, (16,), (1,))
assert_size_stride(primals_3, (4, 4), (4, 1))
assert_size_stride(primals_4, (16, 16), (16, 1))
assert_size_stride(primals_5, (16,), (1,))
assert_size_stride(primals_6, (4, 16), (16, 1))
assert_size_stride(primals_7, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 16), (16, 1), torch.float32)
extern_kernels.mm(primals_3, reinterpret_tensor(primals_1, (4, 16),
(1, 4), 0), out=buf0)
del primals_1
buf1 = buf0
del buf0
get_raw_stream(0)
triton_poi_fused_relu_0[grid(64)](buf1, primals_2, 64, XBLOCK=64,
num_warps=1, num_stages=1)
del primals_2
buf2 = empty_strided_cuda((4, 16), (16, 1), torch.float32)
extern_kernels.mm(buf1, reinterpret_tensor(primals_4, (16, 16), (1,
16), 0), out=buf2)
buf3 = buf2
del buf2
triton_poi_fused_relu_0[grid(64)](buf3, primals_5, 64, XBLOCK=64,
num_warps=1, num_stages=1)
del primals_5
buf4 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.addmm(primals_7, buf3, reinterpret_tensor(primals_6,
(16, 4), (1, 16), 0), alpha=1, beta=1, out=buf4)
del primals_7
buf5 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
triton_poi_fused__softmax_1[grid(16)](buf4, buf5, 16, XBLOCK=16,
num_warps=1, num_stages=1)
buf6 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
triton_poi_fused__softmax_2[grid(16)](buf5, buf6, 16, XBLOCK=16,
num_warps=1, num_stages=1)
buf7 = buf5
del buf5
triton_poi_fused__softmax_2[grid(16)](buf6, buf7, 16, XBLOCK=16,
num_warps=1, num_stages=1)
del buf6
buf8 = torch.ops.aten.multinomial.default(buf7, 1, True)
buf9 = buf8
del buf8
return reinterpret_tensor(buf9, (4,), (1,), 0
), buf7, primals_3, buf1, buf3, buf4, primals_6, primals_4
class A2CActorDiscNew(nn.Module):
def __init__(self, state_dim, action_num):
super().__init__()
self.fc1 = nn.Linear(state_dim, 16)
self.fc2 = nn.Linear(16, 16)
self.fc3 = nn.Linear(16, action_num)
def forward(self, input_0):
primals_1 = self.fc1.weight
primals_2 = self.fc1.bias
primals_4 = self.fc2.weight
primals_5 = self.fc2.bias
primals_6 = self.fc3.weight
primals_7 = self.fc3.bias
primals_3 = input_0
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7])
return output[0], output[1], output[2]
| ikamensh/machin | A2CActorDisc | false | 6,860 | [
"MIT"
] | 1 | af7b423c47bc1412530cf6c96c11bd3af9b3e239 | https://github.com/ikamensh/machin/tree/af7b423c47bc1412530cf6c96c11bd3af9b3e239 | import torch
from torch.distributions import Categorical
import torch as t
import torch.nn as nn
class Model(nn.Module):
def __init__(self, state_dim, action_num):
super().__init__()
self.fc1 = nn.Linear(state_dim, 16)
self.fc2 = nn.Linear(16, 16)
self.fc3 = nn.Linear(16, action_num)
def forward(self, state, action=None):
a = t.relu(self.fc1(state))
a = t.relu(self.fc2(a))
probs = t.softmax(self.fc3(a), dim=1)
dist = Categorical(probs=probs)
act = action if action is not None else dist.sample()
act_entropy = dist.entropy()
act_log_prob = dist.log_prob(act.flatten())
return act, act_log_prob, act_entropy
def get_inputs():
return [torch.rand([4, 4])]
def get_init_inputs():
return [4, 4]
|
MultiHeadedAttention | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/rh/crhy6nilvaajphuuoyup37xl4ncuiyrcb3fnt5aboux6wyvcg7ie.py
# Topologically Sorted Source Nodes: [scores], Original ATen: [aten.clone]
# Source node to ATen node mapping:
# scores => clone
# Graph fragment:
# %clone : [num_users=1] = call_function[target=torch.ops.aten.clone.default](args = (%expand,), kwargs = {memory_format: torch.contiguous_format})
triton_poi_fused_clone_0 = async_compile.triton('triton_poi_fused_clone_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16, 16], tile_hint=TileHint.DEFAULT,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clone_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clone_0(in_ptr0, in_ptr1, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 16
xnumel = 16
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = (yindex // 4)
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + (4*x2) + (64*y1)), xmask & ymask)
tmp1 = tl.load(in_ptr1 + (y0), ymask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(out_ptr0 + (x2 + (16*y3)), tmp2, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/xl/cxldlhjpfliyaeswhsohcdhtqevqxjlvece7kkxd6sy4o7gkfgo3.py
# Topologically Sorted Source Nodes: [scores_2], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# scores_2 => amax, div_1, exp, sub, sum_1
# Graph fragment:
# %amax : [num_users=1] = call_function[target=torch.ops.aten.amax.default](args = (%view_11, [-1], True), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%view_11, %amax), kwargs = {})
# %exp : [num_users=2] = call_function[target=torch.ops.aten.exp.default](args = (%sub,), kwargs = {})
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%exp, [-1], True), kwargs = {})
# %div_1 : [num_users=2] = call_function[target=torch.ops.aten.div.Tensor](args = (%exp, %sum_1), kwargs = {})
triton_per_fused__softmax_1 = async_compile.triton('triton_per_fused__softmax_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[256, 16],
reduction_hint=ReductionHint.INNER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused__softmax_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 2, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused__softmax_1(in_ptr0, out_ptr2, xnumel, rnumel, XBLOCK : tl.constexpr):
xnumel = 256
rnumel = 16
RBLOCK: tl.constexpr = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
roffset = 0
rmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
r1 = rindex
x0 = xindex
tmp0 = tl.load(in_ptr0 + (r1 + (16*x0)), xmask, other=0.0)
tmp1 = tl.broadcast_to(tmp0, [XBLOCK, RBLOCK])
tmp3 = tl.where(xmask, tmp1, float("-inf"))
tmp4 = triton_helpers.max2(tmp3, 1)[:, None]
tmp5 = tmp0 - tmp4
tmp6 = tl_math.exp(tmp5)
tmp7 = tl.broadcast_to(tmp6, [XBLOCK, RBLOCK])
tmp9 = tl.where(xmask, tmp7, 0)
tmp10 = tl.sum(tmp9, 1)[:, None]
tmp11 = tmp6 / tmp10
tl.store(out_ptr2 + (r1 + (16*x0)), tmp11, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/mz/cmzlu2lip25blpsdqeby7ek5757op6xw3pdkxbdediou5szw32tx.py
# Topologically Sorted Source Nodes: [contiguous], Original ATen: [aten.clone]
# Source node to ATen node mapping:
# contiguous => clone_4
# Graph fragment:
# %clone_4 : [num_users=1] = call_function[target=torch.ops.aten.clone.default](args = (%permute_7,), kwargs = {memory_format: torch.contiguous_format})
triton_poi_fused_clone_2 = async_compile.triton('triton_poi_fused_clone_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64, 4], tile_hint=TileHint.SQUARE,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clone_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clone_2(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 64
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 16
y1 = (yindex // 16)
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + (16*x2) + (64*y1)), xmask & ymask, eviction_policy='evict_last')
tl.store(out_ptr0 + (x2 + (4*y3)), tmp0, xmask & ymask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9, primals_10, primals_11 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (4, 4), (4, 1))
assert_size_stride(primals_3, (4, ), (1, ))
assert_size_stride(primals_4, (4, 4), (4, 1))
assert_size_stride(primals_5, (4, ), (1, ))
assert_size_stride(primals_6, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_7, (4, 4), (4, 1))
assert_size_stride(primals_8, (4, ), (1, ))
assert_size_stride(primals_9, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_10, (4, 4), (4, 1))
assert_size_stride(primals_11, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(primals_1, (64, 4), (4, 1), 0), reinterpret_tensor(primals_2, (4, 4), (1, 4), 0), out=buf0)
del primals_2
buf1 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(primals_6, (64, 4), (4, 1), 0), reinterpret_tensor(primals_4, (4, 4), (1, 4), 0), out=buf1)
del primals_4
buf2 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(primals_9, (64, 4), (4, 1), 0), reinterpret_tensor(primals_7, (4, 4), (1, 4), 0), out=buf2)
del primals_7
buf3 = empty_strided_cuda((4, 4, 16, 1), (64, 16, 1, 1), torch.float32)
# Topologically Sorted Source Nodes: [scores], Original ATen: [aten.clone]
stream0 = get_raw_stream(0)
triton_poi_fused_clone_0.run(buf0, primals_3, buf3, 16, 16, grid=grid(16, 16), stream=stream0)
del primals_3
buf4 = reinterpret_tensor(buf0, (4, 4, 1, 16), (64, 16, 16, 1), 0); del buf0 # reuse
# Topologically Sorted Source Nodes: [scores], Original ATen: [aten.clone]
triton_poi_fused_clone_0.run(buf1, primals_5, buf4, 16, 16, grid=grid(16, 16), stream=stream0)
del primals_5
buf5 = empty_strided_cuda((16, 16, 16), (256, 16, 1), torch.float32)
# Topologically Sorted Source Nodes: [scores], Original ATen: [aten.bmm]
extern_kernels.bmm(reinterpret_tensor(buf3, (16, 16, 1), (16, 1, 0), 0), reinterpret_tensor(buf4, (16, 1, 16), (16, 0, 1), 0), out=buf5)
buf8 = empty_strided_cuda((4, 4, 16, 16), (1024, 256, 16, 1), torch.float32)
# Topologically Sorted Source Nodes: [scores_2], Original ATen: [aten._softmax]
triton_per_fused__softmax_1.run(buf5, buf8, 256, 16, grid=grid(256), stream=stream0)
del buf5
buf9 = reinterpret_tensor(buf1, (4, 4, 16, 1), (64, 16, 1, 1), 0); del buf1 # reuse
# Topologically Sorted Source Nodes: [attn_output], Original ATen: [aten.clone]
triton_poi_fused_clone_0.run(buf2, primals_8, buf9, 16, 16, grid=grid(16, 16), stream=stream0)
del primals_8
buf10 = reinterpret_tensor(buf2, (16, 16, 1), (16, 1, 1), 0); del buf2 # reuse
# Topologically Sorted Source Nodes: [attn_output], Original ATen: [aten.bmm]
extern_kernels.bmm(reinterpret_tensor(buf8, (16, 16, 16), (256, 16, 1), 0), reinterpret_tensor(buf9, (16, 16, 1), (16, 1, 0), 0), out=buf10)
buf11 = empty_strided_cuda((4, 16, 4, 1), (64, 4, 1, 1), torch.float32)
# Topologically Sorted Source Nodes: [contiguous], Original ATen: [aten.clone]
triton_poi_fused_clone_2.run(buf10, buf11, 64, 4, grid=grid(64, 4), stream=stream0)
buf12 = reinterpret_tensor(buf10, (64, 4), (4, 1), 0); del buf10 # reuse
# Topologically Sorted Source Nodes: [linear_3], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_11, reinterpret_tensor(buf11, (64, 4), (4, 1), 0), reinterpret_tensor(primals_10, (4, 4), (1, 4), 0), alpha=1, beta=1, out=buf12)
del primals_11
return (reinterpret_tensor(buf12, (4, 16, 4), (64, 4, 1), 0), buf8, reinterpret_tensor(primals_1, (64, 4), (4, 1), 0), reinterpret_tensor(primals_6, (64, 4), (4, 1), 0), reinterpret_tensor(primals_9, (64, 4), (4, 1), 0), buf8, reinterpret_tensor(buf11, (64, 4), (4, 1), 0), primals_10, reinterpret_tensor(buf9, (16, 1, 16), (16, 1, 1), 0), reinterpret_tensor(buf3, (16, 1, 16), (16, 1, 1), 0), reinterpret_tensor(buf4, (16, 16, 1), (16, 1, 16), 0), )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_8 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_9 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_10 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_11 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9, primals_10, primals_11])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import numpy as np
import torch.utils.data
class ScaledDotProductAttention(torch.nn.Module):
"""
Scaled, softmax attention module for Transformer as defined by
Attention(Q, K, V) on pg 4. Returns the final attention vectors as well as
the attention matrices (pairwise scores). """
def __init__(self):
super(ScaledDotProductAttention, self).__init__()
self.softmax = torch.nn.Softmax(dim=-1)
def forward(self, Q, K, V, mask=None, dropout=None):
scores = torch.matmul(Q, K.transpose(-2, -1))
scores = scores / np.sqrt(K.shape[-1])
if mask is not None:
scores = scores.masked_fill(mask == 0, -np.inf)
scores = self.softmax(scores)
if dropout is not None:
scores = dropout(scores)
return torch.matmul(scores, V), scores
class MultiHeadedAttention(torch.nn.Module):
"""
Multi-headed attention layer for the Transformer model. Wraps
ScaledDotProductAttention. Assumes n_heads are applied by splitting up
model in to n_heads, each of size dm / n_heads. Guided by
http://nlp.seas.harvard.edu/2018/04/03/attention.html
"""
def __init__(self, dm, n_heads, dropout=0.1):
super(MultiHeadedAttention, self).__init__()
assert dm % n_heads == 0, 'The dimension of the model must be evenly divisible by the number of attn heads.'
self.dm = dm
self.dk = dm // n_heads
self.n_heads = n_heads
self.wq = torch.nn.Linear(self.dm, self.dm)
self.wk = torch.nn.Linear(self.dm, self.dm)
self.wv = torch.nn.Linear(self.dm, self.dm)
self.wo = torch.nn.Linear(self.dm, self.dm)
self.attn_scores = None
self.attn = ScaledDotProductAttention()
self.dropout = torch.nn.Dropout(dropout)
def forward(self, preQ, preK, preV, mask=None):
n_batch = preQ.shape[0]
Q, K, V = self.wq(preQ), self.wk(preK), self.wv(preV)
Q, K, V = (x.view(n_batch, -1, self.n_heads, self.dk).transpose(1,
2) for x in (Q, K, V))
mask = mask.unsqueeze(1) if mask is not None else None
attn_output, self.attn_scores = self.attn(Q, K, V, mask, self.dropout)
attn_output = attn_output.transpose(1, 2).contiguous().view(n_batch,
-1, self.dm)
return self.wo(attn_output)
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4]), torch.rand(
[4, 4, 4, 4])]
def get_init_inputs():
return [[], {'dm': 4, 'n_heads': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import math as tl_math
import numpy as np
import torch.utils.data
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_clone_0(in_ptr0, in_ptr1, out_ptr0, ynumel, xnumel,
YBLOCK: tl.constexpr, XBLOCK: tl.constexpr):
ynumel = 16
xnumel = 16
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = yindex // 4
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + 4 * x2 + 64 * y1), xmask & ymask)
tmp1 = tl.load(in_ptr1 + y0, ymask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(out_ptr0 + (x2 + 16 * y3), tmp2, xmask & ymask)
@triton.jit
def triton_per_fused__softmax_1(in_ptr0, out_ptr2, xnumel, rnumel, XBLOCK:
tl.constexpr):
xnumel = 256
RBLOCK: tl.constexpr = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
r1 = rindex
x0 = xindex
tmp0 = tl.load(in_ptr0 + (r1 + 16 * x0), xmask, other=0.0)
tmp1 = tl.broadcast_to(tmp0, [XBLOCK, RBLOCK])
tmp3 = tl.where(xmask, tmp1, float('-inf'))
tmp4 = triton_helpers.max2(tmp3, 1)[:, None]
tmp5 = tmp0 - tmp4
tmp6 = tl_math.exp(tmp5)
tmp7 = tl.broadcast_to(tmp6, [XBLOCK, RBLOCK])
tmp9 = tl.where(xmask, tmp7, 0)
tmp10 = tl.sum(tmp9, 1)[:, None]
tmp11 = tmp6 / tmp10
tl.store(out_ptr2 + (r1 + 16 * x0), tmp11, xmask)
@triton.jit
def triton_poi_fused_clone_2(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK: tl.
constexpr, XBLOCK: tl.constexpr):
ynumel = 64
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 16
y1 = yindex // 16
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + 16 * x2 + 64 * y1), xmask & ymask,
eviction_policy='evict_last')
tl.store(out_ptr0 + (x2 + 4 * y3), tmp0, xmask & ymask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7, primals_8, primals_9, primals_10, primals_11) = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (4, 4), (4, 1))
assert_size_stride(primals_3, (4,), (1,))
assert_size_stride(primals_4, (4, 4), (4, 1))
assert_size_stride(primals_5, (4,), (1,))
assert_size_stride(primals_6, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_7, (4, 4), (4, 1))
assert_size_stride(primals_8, (4,), (1,))
assert_size_stride(primals_9, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_10, (4, 4), (4, 1))
assert_size_stride(primals_11, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_1, (64, 4), (4, 1), 0),
reinterpret_tensor(primals_2, (4, 4), (1, 4), 0), out=buf0)
del primals_2
buf1 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_6, (64, 4), (4, 1), 0),
reinterpret_tensor(primals_4, (4, 4), (1, 4), 0), out=buf1)
del primals_4
buf2 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_9, (64, 4), (4, 1), 0),
reinterpret_tensor(primals_7, (4, 4), (1, 4), 0), out=buf2)
del primals_7
buf3 = empty_strided_cuda((4, 4, 16, 1), (64, 16, 1, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_clone_0[grid(16, 16)](buf0, primals_3, buf3, 16,
16, XBLOCK=16, YBLOCK=16, num_warps=4, num_stages=1)
del primals_3
buf4 = reinterpret_tensor(buf0, (4, 4, 1, 16), (64, 16, 16, 1), 0)
del buf0
triton_poi_fused_clone_0[grid(16, 16)](buf1, primals_5, buf4, 16,
16, XBLOCK=16, YBLOCK=16, num_warps=4, num_stages=1)
del primals_5
buf5 = empty_strided_cuda((16, 16, 16), (256, 16, 1), torch.float32)
extern_kernels.bmm(reinterpret_tensor(buf3, (16, 16, 1), (16, 1, 0),
0), reinterpret_tensor(buf4, (16, 1, 16), (16, 0, 1), 0), out=buf5)
buf8 = empty_strided_cuda((4, 4, 16, 16), (1024, 256, 16, 1), torch
.float32)
triton_per_fused__softmax_1[grid(256)](buf5, buf8, 256, 16, XBLOCK=
32, num_warps=4, num_stages=1)
del buf5
buf9 = reinterpret_tensor(buf1, (4, 4, 16, 1), (64, 16, 1, 1), 0)
del buf1
triton_poi_fused_clone_0[grid(16, 16)](buf2, primals_8, buf9, 16,
16, XBLOCK=16, YBLOCK=16, num_warps=4, num_stages=1)
del primals_8
buf10 = reinterpret_tensor(buf2, (16, 16, 1), (16, 1, 1), 0)
del buf2
extern_kernels.bmm(reinterpret_tensor(buf8, (16, 16, 16), (256, 16,
1), 0), reinterpret_tensor(buf9, (16, 16, 1), (16, 1, 0), 0),
out=buf10)
buf11 = empty_strided_cuda((4, 16, 4, 1), (64, 4, 1, 1), torch.float32)
triton_poi_fused_clone_2[grid(64, 4)](buf10, buf11, 64, 4, XBLOCK=4,
YBLOCK=32, num_warps=4, num_stages=1)
buf12 = reinterpret_tensor(buf10, (64, 4), (4, 1), 0)
del buf10
extern_kernels.addmm(primals_11, reinterpret_tensor(buf11, (64, 4),
(4, 1), 0), reinterpret_tensor(primals_10, (4, 4), (1, 4), 0),
alpha=1, beta=1, out=buf12)
del primals_11
return reinterpret_tensor(buf12, (4, 16, 4), (64, 4, 1), 0
), buf8, reinterpret_tensor(primals_1, (64, 4), (4, 1), 0
), reinterpret_tensor(primals_6, (64, 4), (4, 1), 0
), reinterpret_tensor(primals_9, (64, 4), (4, 1), 0
), buf8, reinterpret_tensor(buf11, (64, 4), (4, 1), 0
), primals_10, reinterpret_tensor(buf9, (16, 1, 16), (16, 1, 1), 0
), reinterpret_tensor(buf3, (16, 1, 16), (16, 1, 1), 0
), reinterpret_tensor(buf4, (16, 16, 1), (16, 1, 16), 0)
class ScaledDotProductAttention(torch.nn.Module):
"""
Scaled, softmax attention module for Transformer as defined by
Attention(Q, K, V) on pg 4. Returns the final attention vectors as well as
the attention matrices (pairwise scores). """
def __init__(self):
super(ScaledDotProductAttention, self).__init__()
self.softmax = torch.nn.Softmax(dim=-1)
def forward(self, Q, K, V, mask=None, dropout=None):
scores = torch.matmul(Q, K.transpose(-2, -1))
scores = scores / np.sqrt(K.shape[-1])
if mask is not None:
scores = scores.masked_fill(mask == 0, -np.inf)
scores = self.softmax(scores)
if dropout is not None:
scores = dropout(scores)
return torch.matmul(scores, V), scores
class MultiHeadedAttentionNew(torch.nn.Module):
"""
Multi-headed attention layer for the Transformer model. Wraps
ScaledDotProductAttention. Assumes n_heads are applied by splitting up
model in to n_heads, each of size dm / n_heads. Guided by
http://nlp.seas.harvard.edu/2018/04/03/attention.html
"""
def __init__(self, dm, n_heads, dropout=0.1):
super(MultiHeadedAttentionNew, self).__init__()
assert dm % n_heads == 0, 'The dimension of the model must be evenly divisible by the number of attn heads.'
self.dm = dm
self.dk = dm // n_heads
self.n_heads = n_heads
self.wq = torch.nn.Linear(self.dm, self.dm)
self.wk = torch.nn.Linear(self.dm, self.dm)
self.wv = torch.nn.Linear(self.dm, self.dm)
self.wo = torch.nn.Linear(self.dm, self.dm)
self.attn_scores = None
self.attn = ScaledDotProductAttention()
self.dropout = torch.nn.Dropout(dropout)
def forward(self, input_0, input_1, input_2):
primals_2 = self.wq.weight
primals_3 = self.wq.bias
primals_4 = self.wk.weight
primals_5 = self.wk.bias
primals_7 = self.wv.weight
primals_8 = self.wv.bias
primals_10 = self.wo.weight
primals_11 = self.wo.bias
primals_1 = input_0
primals_6 = input_1
primals_9 = input_2
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7, primals_8, primals_9,
primals_10, primals_11])
return output[0]
| icemansina/protein-transformer | MultiHeadedAttention | false | 6,861 | [
"BSD-3-Clause"
] | 1 | 4e73b17f2a4b89ba1a9f6703976d1a31b7a8a5eb | https://github.com/icemansina/protein-transformer/tree/4e73b17f2a4b89ba1a9f6703976d1a31b7a8a5eb | import torch
import numpy as np
import torch.utils.data
class ScaledDotProductAttention(torch.nn.Module):
"""
Scaled, softmax attention module for Transformer as defined by
Attention(Q, K, V) on pg 4. Returns the final attention vectors as well as
the attention matrices (pairwise scores). """
def __init__(self):
super().__init__()
self.softmax = torch.nn.Softmax(dim=-1)
def forward(self, Q, K, V, mask=None, dropout=None):
scores = torch.matmul(Q, K.transpose(-2, -1))
scores = scores / np.sqrt(K.shape[-1])
if mask is not None:
scores = scores.masked_fill(mask == 0, -np.inf)
scores = self.softmax(scores)
if dropout is not None:
scores = dropout(scores)
return torch.matmul(scores, V), scores
class Model(torch.nn.Module):
"""
Multi-headed attention layer for the Transformer model. Wraps
ScaledDotProductAttention. Assumes n_heads are applied by splitting up
model in to n_heads, each of size dm / n_heads. Guided by
http://nlp.seas.harvard.edu/2018/04/03/attention.html
"""
def __init__(self, dm, n_heads, dropout=0.1):
super().__init__()
assert dm % n_heads == 0, 'The dimension of the model must be evenly divisible by the number of attn heads.'
self.dm = dm
self.dk = dm // n_heads
self.n_heads = n_heads
self.wq = torch.nn.Linear(self.dm, self.dm)
self.wk = torch.nn.Linear(self.dm, self.dm)
self.wv = torch.nn.Linear(self.dm, self.dm)
self.wo = torch.nn.Linear(self.dm, self.dm)
self.attn_scores = None
self.attn = ScaledDotProductAttention()
self.dropout = torch.nn.Dropout(dropout)
def forward(self, preQ, preK, preV, mask=None):
n_batch = preQ.shape[0]
Q, K, V = self.wq(preQ), self.wk(preK), self.wv(preV)
Q, K, V = (x.view(n_batch, -1, self.n_heads, self.dk).transpose(1,
2) for x in (Q, K, V))
mask = mask.unsqueeze(1) if mask is not None else None
attn_output, self.attn_scores = self.attn(Q, K, V, mask, self.dropout)
attn_output = attn_output.transpose(1, 2).contiguous().view(n_batch,
-1, self.dm)
return self.wo(attn_output)
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4]), torch.rand(
[4, 4, 4, 4])]
def get_init_inputs():
return [4, 4]
|
QNet | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/v7/cv7zazascu4rpkkwoxbiwk6c2le2e6wshdhae73bmaoapelvwguv.py
# Topologically Sorted Source Nodes: [a], Original ATen: [aten.relu, aten.threshold_backward]
# Source node to ATen node mapping:
# a => relu
# Graph fragment:
# %relu : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%view_1,), kwargs = {})
# %le_1 : [num_users=1] = call_function[target=torch.ops.aten.le.Scalar](args = (%relu, 0), kwargs = {})
triton_poi_fused_relu_threshold_backward_0 = async_compile.triton('triton_poi_fused_relu_threshold_backward_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[1024],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*i1', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_relu_threshold_backward_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_relu_threshold_backward_0(in_out_ptr0, in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 1024
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 16
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(in_out_ptr0 + (x2), tmp4, xmask)
tl.store(out_ptr0 + (x2), tmp6, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/xf/cxf4na4u2jaehhvves35mikpqe7m3jl2pikqdmab7ftwzhsdy3qk.py
# Topologically Sorted Source Nodes: [softmax], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# softmax => amax, div, exp, sub, sum_1
# Graph fragment:
# %amax : [num_users=1] = call_function[target=torch.ops.aten.amax.default](args = (%view_6, [-1], True), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%view_6, %amax), kwargs = {})
# %exp : [num_users=2] = call_function[target=torch.ops.aten.exp.default](args = (%sub,), kwargs = {})
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%exp, [-1], True), kwargs = {})
# %div : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%exp, %sum_1), kwargs = {})
triton_per_fused__softmax_1 = async_compile.triton('triton_per_fused__softmax_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[256, 16],
reduction_hint=ReductionHint.INNER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused__softmax_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 2, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused__softmax_1(in_ptr0, out_ptr2, xnumel, rnumel, XBLOCK : tl.constexpr):
xnumel = 256
rnumel = 10
RBLOCK: tl.constexpr = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
roffset = 0
rmask = rindex < rnumel
r1 = rindex
x0 = xindex
tmp0 = tl.load(in_ptr0 + (r1 + (10*x0)), rmask & xmask, other=0.0)
tmp1 = tl.broadcast_to(tmp0, [XBLOCK, RBLOCK])
tmp3 = tl.where(rmask & xmask, tmp1, float("-inf"))
tmp4 = triton_helpers.max2(tmp3, 1)[:, None]
tmp5 = tmp0 - tmp4
tmp6 = tl_math.exp(tmp5)
tmp7 = tl.broadcast_to(tmp6, [XBLOCK, RBLOCK])
tmp9 = tl.where(rmask & xmask, tmp7, 0)
tmp10 = tl.sum(tmp9, 1)[:, None]
tmp11 = tmp6 / tmp10
tl.store(out_ptr2 + (r1 + (10*x0)), tmp11, rmask & xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7 = args
args.clear()
assert_size_stride(primals_1, (16, 4), (4, 1))
assert_size_stride(primals_2, (16, ), (1, ))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_4, (16, 16), (16, 1))
assert_size_stride(primals_5, (16, ), (1, ))
assert_size_stride(primals_6, (40, 16), (16, 1))
assert_size_stride(primals_7, (40, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 16), (16, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(primals_3, (64, 4), (4, 1), 0), reinterpret_tensor(primals_1, (4, 16), (1, 4), 0), out=buf0)
del primals_1
buf1 = reinterpret_tensor(buf0, (4, 4, 4, 16), (256, 64, 16, 1), 0); del buf0 # reuse
buf9 = empty_strided_cuda((4, 4, 4, 16), (256, 64, 16, 1), torch.bool)
# Topologically Sorted Source Nodes: [a], Original ATen: [aten.relu, aten.threshold_backward]
stream0 = get_raw_stream(0)
triton_poi_fused_relu_threshold_backward_0.run(buf1, primals_2, buf9, 1024, grid=grid(1024), stream=stream0)
del primals_2
buf2 = empty_strided_cuda((64, 16), (16, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(buf1, (64, 16), (16, 1), 0), reinterpret_tensor(primals_4, (16, 16), (1, 16), 0), out=buf2)
buf3 = reinterpret_tensor(buf2, (4, 4, 4, 16), (256, 64, 16, 1), 0); del buf2 # reuse
buf8 = empty_strided_cuda((4, 4, 4, 16), (256, 64, 16, 1), torch.bool)
# Topologically Sorted Source Nodes: [a_1], Original ATen: [aten.relu, aten.threshold_backward]
triton_poi_fused_relu_threshold_backward_0.run(buf3, primals_5, buf8, 1024, grid=grid(1024), stream=stream0)
del primals_5
buf4 = empty_strided_cuda((64, 40), (40, 1), torch.float32)
# Topologically Sorted Source Nodes: [linear_2], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_7, reinterpret_tensor(buf3, (64, 16), (16, 1), 0), reinterpret_tensor(primals_6, (16, 40), (1, 16), 0), alpha=1, beta=1, out=buf4)
del primals_7
buf7 = empty_strided_cuda((64, 4, 10), (40, 10, 1), torch.float32)
# Topologically Sorted Source Nodes: [softmax], Original ATen: [aten._softmax]
triton_per_fused__softmax_1.run(buf4, buf7, 256, 10, grid=grid(256), stream=stream0)
del buf4
return (buf7, reinterpret_tensor(primals_3, (64, 4), (4, 1), 0), reinterpret_tensor(buf1, (64, 16), (16, 1), 0), reinterpret_tensor(buf3, (64, 16), (16, 1), 0), buf7, primals_6, buf8, primals_4, buf9, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((16, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((16, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((16, 16), (16, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((16, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((40, 16), (16, 1), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((40, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch as t
import torch.nn as nn
class QNet(nn.Module):
def __init__(self, state_dim, action_num, atom_num=10):
super().__init__()
self.fc1 = nn.Linear(state_dim, 16)
self.fc2 = nn.Linear(16, 16)
self.fc3 = nn.Linear(16, action_num * atom_num)
self.action_num = action_num
self.atom_num = atom_num
def forward(self, state):
a = t.relu(self.fc1(state))
a = t.relu(self.fc2(a))
return t.softmax(self.fc3(a).view(-1, self.action_num, self.
atom_num), dim=-1)
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'state_dim': 4, 'action_num': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import math as tl_math
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_relu_threshold_backward_0(in_out_ptr0, in_ptr0,
out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 1024
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 16
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(in_out_ptr0 + x2, tmp4, xmask)
tl.store(out_ptr0 + x2, tmp6, xmask)
@triton.jit
def triton_per_fused__softmax_1(in_ptr0, out_ptr2, xnumel, rnumel, XBLOCK:
tl.constexpr):
xnumel = 256
rnumel = 10
RBLOCK: tl.constexpr = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
rmask = rindex < rnumel
r1 = rindex
x0 = xindex
tmp0 = tl.load(in_ptr0 + (r1 + 10 * x0), rmask & xmask, other=0.0)
tmp1 = tl.broadcast_to(tmp0, [XBLOCK, RBLOCK])
tmp3 = tl.where(rmask & xmask, tmp1, float('-inf'))
tmp4 = triton_helpers.max2(tmp3, 1)[:, None]
tmp5 = tmp0 - tmp4
tmp6 = tl_math.exp(tmp5)
tmp7 = tl.broadcast_to(tmp6, [XBLOCK, RBLOCK])
tmp9 = tl.where(rmask & xmask, tmp7, 0)
tmp10 = tl.sum(tmp9, 1)[:, None]
tmp11 = tmp6 / tmp10
tl.store(out_ptr2 + (r1 + 10 * x0), tmp11, rmask & xmask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7) = args
args.clear()
assert_size_stride(primals_1, (16, 4), (4, 1))
assert_size_stride(primals_2, (16,), (1,))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_4, (16, 16), (16, 1))
assert_size_stride(primals_5, (16,), (1,))
assert_size_stride(primals_6, (40, 16), (16, 1))
assert_size_stride(primals_7, (40,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 16), (16, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_3, (64, 4), (4, 1), 0),
reinterpret_tensor(primals_1, (4, 16), (1, 4), 0), out=buf0)
del primals_1
buf1 = reinterpret_tensor(buf0, (4, 4, 4, 16), (256, 64, 16, 1), 0)
del buf0
buf9 = empty_strided_cuda((4, 4, 4, 16), (256, 64, 16, 1), torch.bool)
get_raw_stream(0)
triton_poi_fused_relu_threshold_backward_0[grid(1024)](buf1,
primals_2, buf9, 1024, XBLOCK=128, num_warps=4, num_stages=1)
del primals_2
buf2 = empty_strided_cuda((64, 16), (16, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(buf1, (64, 16), (16, 1), 0),
reinterpret_tensor(primals_4, (16, 16), (1, 16), 0), out=buf2)
buf3 = reinterpret_tensor(buf2, (4, 4, 4, 16), (256, 64, 16, 1), 0)
del buf2
buf8 = empty_strided_cuda((4, 4, 4, 16), (256, 64, 16, 1), torch.bool)
triton_poi_fused_relu_threshold_backward_0[grid(1024)](buf3,
primals_5, buf8, 1024, XBLOCK=128, num_warps=4, num_stages=1)
del primals_5
buf4 = empty_strided_cuda((64, 40), (40, 1), torch.float32)
extern_kernels.addmm(primals_7, reinterpret_tensor(buf3, (64, 16),
(16, 1), 0), reinterpret_tensor(primals_6, (16, 40), (1, 16), 0
), alpha=1, beta=1, out=buf4)
del primals_7
buf7 = empty_strided_cuda((64, 4, 10), (40, 10, 1), torch.float32)
triton_per_fused__softmax_1[grid(256)](buf4, buf7, 256, 10, XBLOCK=
8, num_warps=2, num_stages=1)
del buf4
return buf7, reinterpret_tensor(primals_3, (64, 4), (4, 1), 0
), reinterpret_tensor(buf1, (64, 16), (16, 1), 0), reinterpret_tensor(
buf3, (64, 16), (16, 1), 0), buf7, primals_6, buf8, primals_4, buf9
class QNetNew(nn.Module):
def __init__(self, state_dim, action_num, atom_num=10):
super().__init__()
self.fc1 = nn.Linear(state_dim, 16)
self.fc2 = nn.Linear(16, 16)
self.fc3 = nn.Linear(16, action_num * atom_num)
self.action_num = action_num
self.atom_num = atom_num
def forward(self, input_0):
primals_1 = self.fc1.weight
primals_2 = self.fc1.bias
primals_4 = self.fc2.weight
primals_5 = self.fc2.bias
primals_6 = self.fc3.weight
primals_7 = self.fc3.bias
primals_3 = input_0
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7])
return output[0]
| ikamensh/machin | QNet | false | 6,862 | [
"MIT"
] | 1 | af7b423c47bc1412530cf6c96c11bd3af9b3e239 | https://github.com/ikamensh/machin/tree/af7b423c47bc1412530cf6c96c11bd3af9b3e239 | import torch
import torch as t
import torch.nn as nn
class Model(nn.Module):
def __init__(self, state_dim, action_num, atom_num=10):
super().__init__()
self.fc1 = nn.Linear(state_dim, 16)
self.fc2 = nn.Linear(16, 16)
self.fc3 = nn.Linear(16, action_num * atom_num)
self.action_num = action_num
self.atom_num = atom_num
def forward(self, state):
a = t.relu(self.fc1(state))
a = t.relu(self.fc2(a))
return t.softmax(self.fc3(a).view(-1, self.action_num, self.
atom_num), dim=-1)
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4, 4]
|
CosineBasisLinear | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/vm/cvmxb4fm3754gb5xbvn445attxbnlnqmgxmnql4g235hqqjxgk4t.py
# Topologically Sorted Source Nodes: [arange, i_pi, mul_1, embedding], Original ATen: [aten.arange, aten.mul, aten.cos]
# Source node to ATen node mapping:
# arange => add, convert_element_type, iota, mul
# embedding => cos
# i_pi => mul_1
# mul_1 => mul_2
# Graph fragment:
# %iota : [num_users=1] = call_function[target=torch.ops.prims.iota.default](args = (4,), kwargs = {start: 0, step: 1, dtype: torch.int64, device: cuda:0, requires_grad: False})
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%iota, 1), kwargs = {})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul, 1), kwargs = {})
# %convert_element_type : [num_users=1] = call_function[target=torch.ops.prims.convert_element_type.default](args = (%add, torch.float32), kwargs = {})
# %mul_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%convert_element_type, 3.141592653589793), kwargs = {})
# %mul_2 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%unsqueeze, %mul_1), kwargs = {})
# %cos : [num_users=1] = call_function[target=torch.ops.aten.cos.default](args = (%mul_2,), kwargs = {})
triton_poi_fused_arange_cos_mul_0 = async_compile.triton('triton_poi_fused_arange_cos_mul_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[1024],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_arange_cos_mul_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_arange_cos_mul_0(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 1024
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = (xindex // 4)
x0 = xindex % 4
x2 = xindex
tmp0 = tl.load(in_ptr0 + (x1), xmask, eviction_policy='evict_last')
tmp1 = 1 + x0
tmp2 = tmp1.to(tl.float32)
tmp3 = 3.141592653589793
tmp4 = tmp2 * tmp3
tmp5 = tmp0 * tmp4
tmp6 = tl_math.cos(tmp5)
tl.store(out_ptr0 + (x2), tmp6, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (4, 4), (4, 1))
assert_size_stride(primals_3, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4, 4, 4), (256, 64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [arange, i_pi, mul_1, embedding], Original ATen: [aten.arange, aten.mul, aten.cos]
stream0 = get_raw_stream(0)
triton_poi_fused_arange_cos_mul_0.run(primals_1, buf0, 1024, grid=grid(1024), stream=stream0)
del primals_1
buf1 = empty_strided_cuda((256, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [out], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_3, reinterpret_tensor(buf0, (256, 4), (4, 1), 0), reinterpret_tensor(primals_2, (4, 4), (1, 4), 0), alpha=1, beta=1, out=buf1)
del primals_2
del primals_3
return (reinterpret_tensor(buf1, (4, 4, 4, 4, 4), (256, 64, 16, 4, 1), 0), reinterpret_tensor(buf0, (256, 4), (4, 1), 0), )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import numpy as np
import torch.nn as nn
def cosine_basis_functions(x, n_basis_functions=64):
"""Cosine basis functions used to embed quantile thresholds.
Args:
x (torch.Tensor): Input.
n_basis_functions (int): Number of cosine basis functions.
Returns:
ndarray: Embedding with shape of (x.shape + (n_basis_functions,)).
"""
i_pi = torch.arange(1, n_basis_functions + 1, dtype=torch.float, device
=x.device) * np.pi
embedding = torch.cos(x[..., None] * i_pi)
assert embedding.shape == x.shape + (n_basis_functions,)
return embedding
class CosineBasisLinear(nn.Module):
"""Linear layer following cosine basis functions.
Args:
n_basis_functions (int): Number of cosine basis functions.
out_size (int): Output size.
"""
def __init__(self, n_basis_functions, out_size):
super().__init__()
self.linear = nn.Linear(n_basis_functions, out_size)
self.n_basis_functions = n_basis_functions
self.out_size = out_size
def forward(self, x):
"""Evaluate.
Args:
x (torch.Tensor): Input.
Returns:
torch.Tensor: Output with shape of (x.shape + (out_size,)).
"""
h = cosine_basis_functions(x, self.n_basis_functions)
h = h.reshape(-1, self.n_basis_functions)
out = self.linear(h)
out = out.reshape(*x.shape, self.out_size)
return out
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'n_basis_functions': 4, 'out_size': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime.triton_helpers import math as tl_math
import numpy as np
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_arange_cos_mul_0(in_ptr0, out_ptr0, xnumel, XBLOCK: tl
.constexpr):
xnumel = 1024
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = xindex // 4
x0 = xindex % 4
x2 = xindex
tmp0 = tl.load(in_ptr0 + x1, xmask, eviction_policy='evict_last')
tmp1 = 1 + x0
tmp2 = tmp1.to(tl.float32)
tmp3 = 3.141592653589793
tmp4 = tmp2 * tmp3
tmp5 = tmp0 * tmp4
tmp6 = tl_math.cos(tmp5)
tl.store(out_ptr0 + x2, tmp6, xmask)
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (4, 4), (4, 1))
assert_size_stride(primals_3, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4, 4, 4), (256, 64, 16, 4, 1),
torch.float32)
get_raw_stream(0)
triton_poi_fused_arange_cos_mul_0[grid(1024)](primals_1, buf0, 1024,
XBLOCK=256, num_warps=4, num_stages=1)
del primals_1
buf1 = empty_strided_cuda((256, 4), (4, 1), torch.float32)
extern_kernels.addmm(primals_3, reinterpret_tensor(buf0, (256, 4),
(4, 1), 0), reinterpret_tensor(primals_2, (4, 4), (1, 4), 0),
alpha=1, beta=1, out=buf1)
del primals_2
del primals_3
return reinterpret_tensor(buf1, (4, 4, 4, 4, 4), (256, 64, 16, 4, 1), 0
), reinterpret_tensor(buf0, (256, 4), (4, 1), 0)
def cosine_basis_functions(x, n_basis_functions=64):
"""Cosine basis functions used to embed quantile thresholds.
Args:
x (torch.Tensor): Input.
n_basis_functions (int): Number of cosine basis functions.
Returns:
ndarray: Embedding with shape of (x.shape + (n_basis_functions,)).
"""
i_pi = torch.arange(1, n_basis_functions + 1, dtype=torch.float, device
=x.device) * np.pi
embedding = torch.cos(x[..., None] * i_pi)
assert embedding.shape == x.shape + (n_basis_functions,)
return embedding
class CosineBasisLinearNew(nn.Module):
"""Linear layer following cosine basis functions.
Args:
n_basis_functions (int): Number of cosine basis functions.
out_size (int): Output size.
"""
def __init__(self, n_basis_functions, out_size):
super().__init__()
self.linear = nn.Linear(n_basis_functions, out_size)
self.n_basis_functions = n_basis_functions
self.out_size = out_size
def forward(self, input_0):
primals_2 = self.linear.weight
primals_3 = self.linear.bias
primals_1 = input_0
output = call([primals_1, primals_2, primals_3])
return output[0]
| imatge-upc/pixelcoordEDL | CosineBasisLinear | false | 6,863 | [
"MIT"
] | 1 | 353632feed6ac8c93758c1a2a1b7a477e7ff053c | https://github.com/imatge-upc/pixelcoordEDL/tree/353632feed6ac8c93758c1a2a1b7a477e7ff053c | import torch
import numpy as np
import torch.nn as nn
def cosine_basis_functions(x, n_basis_functions=64):
"""Cosine basis functions used to embed quantile thresholds.
Args:
x (torch.Tensor): Input.
n_basis_functions (int): Number of cosine basis functions.
Returns:
ndarray: Embedding with shape of (x.shape + (n_basis_functions,)).
"""
i_pi = torch.arange(1, n_basis_functions + 1, dtype=torch.float, device
=x.device) * np.pi
embedding = torch.cos(x[..., None] * i_pi)
assert embedding.shape == x.shape + (n_basis_functions,)
return embedding
class Model(nn.Module):
"""Linear layer following cosine basis functions.
Args:
n_basis_functions (int): Number of cosine basis functions.
out_size (int): Output size.
"""
def __init__(self, n_basis_functions, out_size):
super().__init__()
self.linear = nn.Linear(n_basis_functions, out_size)
self.n_basis_functions = n_basis_functions
self.out_size = out_size
def forward(self, x):
"""Evaluate.
Args:
x (torch.Tensor): Input.
Returns:
torch.Tensor: Output with shape of (x.shape + (out_size,)).
"""
h = cosine_basis_functions(x, self.n_basis_functions)
h = h.reshape(-1, self.n_basis_functions)
out = self.linear(h)
out = out.reshape(*x.shape, self.out_size)
return out
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4, 4]
|
GatedActivation | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/nj/cnjk6lhqhjposltiibohjqghrimpe65rvwj3ht3r7foss23xnfod.py
# Topologically Sorted Source Nodes: [tanh, sigmoid, mul], Original ATen: [aten.tanh, aten.sigmoid, aten.mul]
# Source node to ATen node mapping:
# mul => mul
# sigmoid => sigmoid
# tanh => tanh
# Graph fragment:
# %tanh : [num_users=1] = call_function[target=torch.ops.aten.tanh.default](args = (%getitem,), kwargs = {})
# %sigmoid : [num_users=1] = call_function[target=torch.ops.aten.sigmoid.default](args = (%getitem_1,), kwargs = {})
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%tanh, %sigmoid), kwargs = {})
triton_poi_fused_mul_sigmoid_tanh_0 = async_compile.triton('triton_poi_fused_mul_sigmoid_tanh_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[128],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_mul_sigmoid_tanh_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_mul_sigmoid_tanh_0(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 128
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 32
x1 = (xindex // 32)
x2 = xindex
tmp0 = tl.load(in_ptr0 + (x0 + (64*x1)), xmask)
tmp2 = tl.load(in_ptr0 + (32 + x0 + (64*x1)), xmask)
tmp1 = libdevice.tanh(tmp0)
tmp3 = tl.sigmoid(tmp2)
tmp4 = tmp1 * tmp3
tl.store(out_ptr0 + (x2), tmp4, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 2, 4, 4), (32, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [tanh, sigmoid, mul], Original ATen: [aten.tanh, aten.sigmoid, aten.mul]
stream0 = get_raw_stream(0)
triton_poi_fused_mul_sigmoid_tanh_0.run(arg0_1, buf0, 128, grid=grid(128), stream=stream0)
del arg0_1
return (buf0, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([arg0_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.nn.functional as F
class GatedActivation(nn.Module):
def __init__(self):
super().__init__()
def forward(self, x):
x, y = x.chunk(2, dim=1)
return F.tanh(x) * F.sigmoid(y)
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {}]
| import torch
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime.triton_helpers import libdevice
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_poi_fused_mul_sigmoid_tanh_0(in_ptr0, out_ptr0, xnumel, XBLOCK:
tl.constexpr):
xnumel = 128
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 32
x1 = xindex // 32
x2 = xindex
tmp0 = tl.load(in_ptr0 + (x0 + 64 * x1), xmask)
tmp2 = tl.load(in_ptr0 + (32 + x0 + 64 * x1), xmask)
tmp1 = libdevice.tanh(tmp0)
tmp3 = tl.sigmoid(tmp2)
tmp4 = tmp1 * tmp3
tl.store(out_ptr0 + x2, tmp4, xmask)
def call(args):
arg0_1, = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 2, 4, 4), (32, 16, 4, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_mul_sigmoid_tanh_0[grid(128)](arg0_1, buf0, 128,
XBLOCK=128, num_warps=4, num_stages=1)
del arg0_1
return buf0,
class GatedActivationNew(nn.Module):
def __init__(self):
super().__init__()
def forward(self, input_0):
arg0_1 = input_0
output = call([arg0_1])
return output[0]
| imatge-upc/pixelcoordEDL | GatedActivation | false | 6,864 | [
"MIT"
] | 1 | 353632feed6ac8c93758c1a2a1b7a477e7ff053c | https://github.com/imatge-upc/pixelcoordEDL/tree/353632feed6ac8c93758c1a2a1b7a477e7ff053c | import torch
import torch.nn as nn
import torch.nn.functional as F
class Model(nn.Module):
def __init__(self):
super().__init__()
def forward(self, x):
x, y = x.chunk(2, dim=1)
return F.tanh(x) * F.sigmoid(y)
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return []
|
A2CActorCont | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/v7/cv7zazascu4rpkkwoxbiwk6c2le2e6wshdhae73bmaoapelvwguv.py
# Topologically Sorted Source Nodes: [a], Original ATen: [aten.relu, aten.threshold_backward]
# Source node to ATen node mapping:
# a => relu
# Graph fragment:
# %relu : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%view_1,), kwargs = {})
# %le_1 : [num_users=1] = call_function[target=torch.ops.aten.le.Scalar](args = (%relu, 0), kwargs = {})
triton_poi_fused_relu_threshold_backward_0 = async_compile.triton('triton_poi_fused_relu_threshold_backward_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[1024],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*i1', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_relu_threshold_backward_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_relu_threshold_backward_0(in_out_ptr0, in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 1024
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 16
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(in_out_ptr0 + (x2), tmp4, xmask)
tl.store(out_ptr0 + (x2), tmp6, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/oz/coz2vvm7j43dgp2scg63a7xyihnzn7xcwkbiwytoo3i3naiahfwp.py
# Topologically Sorted Source Nodes: [tanh, mu], Original ATen: [aten.tanh, aten.mul]
# Source node to ATen node mapping:
# mu => mul
# tanh => tanh
# Graph fragment:
# %tanh : [num_users=1] = call_function[target=torch.ops.aten.tanh.default](args = (%view_5,), kwargs = {})
# %mul : [num_users=2] = call_function[target=torch.ops.aten.mul.Tensor](args = (%tanh, 2.0), kwargs = {})
triton_poi_fused_mul_tanh_1 = async_compile.triton('triton_poi_fused_mul_tanh_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_mul_tanh_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_mul_tanh_1(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (x0), xmask)
tmp1 = libdevice.tanh(tmp0)
tmp2 = 2.0
tmp3 = tmp1 * tmp2
tl.store(out_ptr0 + (x0), tmp3, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/vn/cvnduo4s466rgfr5qur2ih4vzvlqu5mjrir76fxacpwhzg7t533r.py
# Topologically Sorted Source Nodes: [sigma, log, action_entropy], Original ATen: [aten.softplus, aten.log, aten.add]
# Source node to ATen node mapping:
# action_entropy => add
# log => log
# sigma => exp, gt, log1p, where
# Graph fragment:
# %exp : [num_users=1] = call_function[target=torch.ops.aten.exp.default](args = (%view_7,), kwargs = {})
# %log1p : [num_users=1] = call_function[target=torch.ops.aten.log1p.default](args = (%exp,), kwargs = {})
# %gt : [num_users=1] = call_function[target=torch.ops.aten.gt.Scalar](args = (%view_7, 20), kwargs = {})
# %where : [num_users=3] = call_function[target=torch.ops.aten.where.self](args = (%gt, %view_7, %log1p), kwargs = {})
# %log : [num_users=2] = call_function[target=torch.ops.aten.log.default](args = (%where,), kwargs = {})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%log, 1.4189385332046727), kwargs = {})
triton_poi_fused_add_log_softplus_2 = async_compile.triton('triton_poi_fused_add_log_softplus_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_log_softplus_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_log_softplus_2(in_ptr0, out_ptr0, out_ptr1, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (x0), xmask)
tmp1 = 20.0
tmp2 = tmp0 > tmp1
tmp3 = tl_math.exp(tmp0)
tmp4 = libdevice.log1p(tmp3)
tmp5 = tl.where(tmp2, tmp0, tmp4)
tmp6 = tl_math.log(tmp5)
tmp7 = 1.4189385332046727
tmp8 = tmp6 + tmp7
tl.store(out_ptr0 + (x0), tmp5, xmask)
tl.store(out_ptr1 + (x0), tmp8, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/5x/c5xxopo7kvrfl7zh4iho33rpirl2hnwrmhemgvjglcq7m3e6yekh.py
# Topologically Sorted Source Nodes: [log, action_1, var, sub, pow_2, neg, mul_1, truediv, sub_1, action_log_prob], Original ATen: [aten.log, aten.clamp, aten.pow, aten.sub, aten.neg, aten.mul, aten.div]
# Source node to ATen node mapping:
# action_1 => clamp_max, clamp_min
# action_log_prob => sub_2
# log => log
# mul_1 => mul_1
# neg => neg
# pow_2 => pow_2
# sub => sub
# sub_1 => sub_1
# truediv => div
# var => pow_1
# Graph fragment:
# %log : [num_users=2] = call_function[target=torch.ops.aten.log.default](args = (%where,), kwargs = {})
# %clamp_min : [num_users=1] = call_function[target=torch.ops.aten.clamp_min.default](args = (%normal, -4), kwargs = {})
# %clamp_max : [num_users=2] = call_function[target=torch.ops.aten.clamp_max.default](args = (%clamp_min, 4), kwargs = {})
# %pow_1 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%where, 2), kwargs = {})
# %sub : [num_users=2] = call_function[target=torch.ops.aten.sub.Tensor](args = (%clamp_max, %mul), kwargs = {})
# %pow_2 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%sub, 2), kwargs = {})
# %neg : [num_users=1] = call_function[target=torch.ops.aten.neg.default](args = (%pow_2,), kwargs = {})
# %mul_1 : [num_users=2] = call_function[target=torch.ops.aten.mul.Tensor](args = (%pow_1, 2), kwargs = {})
# %div : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%neg, %mul_1), kwargs = {})
# %sub_1 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%div, %log), kwargs = {})
# %sub_2 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%sub_1, 0.9189385332046727), kwargs = {})
triton_poi_fused_clamp_div_log_mul_neg_pow_sub_3 = async_compile.triton('triton_poi_fused_clamp_div_log_mul_neg_pow_sub_3', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: '*fp32', 5: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clamp_div_log_mul_neg_pow_sub_3', 'mutated_arg_names': ['in_out_ptr0', 'in_out_ptr1'], 'no_x_dim': False, 'num_load': 3, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clamp_div_log_mul_neg_pow_sub_3(in_out_ptr0, in_out_ptr1, in_ptr0, out_ptr0, out_ptr1, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_out_ptr0 + (x0), xmask)
tmp5 = tl.load(in_out_ptr1 + (x0), xmask)
tmp7 = tl.load(in_ptr0 + (x0), xmask)
tmp1 = -4.0
tmp2 = triton_helpers.maximum(tmp0, tmp1)
tmp3 = 4.0
tmp4 = triton_helpers.minimum(tmp2, tmp3)
tmp6 = tmp4 - tmp5
tmp8 = tmp7 * tmp7
tmp9 = 2.0
tmp10 = tmp8 * tmp9
tmp11 = tmp6 * tmp6
tmp12 = -tmp11
tmp13 = tmp12 / tmp10
tmp14 = tl_math.log(tmp7)
tmp15 = tmp13 - tmp14
tmp16 = 0.9189385332046727
tmp17 = tmp15 - tmp16
tl.store(in_out_ptr0 + (x0), tmp4, xmask)
tl.store(in_out_ptr1 + (x0), tmp6, xmask)
tl.store(out_ptr0 + (x0), tmp10, xmask)
tl.store(out_ptr1 + (x0), tmp17, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9 = args
args.clear()
assert_size_stride(primals_1, (16, 4), (4, 1))
assert_size_stride(primals_2, (16, ), (1, ))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_4, (16, 16), (16, 1))
assert_size_stride(primals_5, (16, ), (1, ))
assert_size_stride(primals_6, (4, 16), (16, 1))
assert_size_stride(primals_7, (4, ), (1, ))
assert_size_stride(primals_8, (4, 16), (16, 1))
assert_size_stride(primals_9, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 16), (16, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(primals_3, (64, 4), (4, 1), 0), reinterpret_tensor(primals_1, (4, 16), (1, 4), 0), out=buf0)
del primals_1
buf1 = reinterpret_tensor(buf0, (4, 4, 4, 16), (256, 64, 16, 1), 0); del buf0 # reuse
buf16 = empty_strided_cuda((4, 4, 4, 16), (256, 64, 16, 1), torch.bool)
# Topologically Sorted Source Nodes: [a], Original ATen: [aten.relu, aten.threshold_backward]
stream0 = get_raw_stream(0)
triton_poi_fused_relu_threshold_backward_0.run(buf1, primals_2, buf16, 1024, grid=grid(1024), stream=stream0)
del primals_2
buf2 = empty_strided_cuda((64, 16), (16, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(buf1, (64, 16), (16, 1), 0), reinterpret_tensor(primals_4, (16, 16), (1, 16), 0), out=buf2)
buf3 = reinterpret_tensor(buf2, (4, 4, 4, 16), (256, 64, 16, 1), 0); del buf2 # reuse
buf15 = empty_strided_cuda((4, 4, 4, 16), (256, 64, 16, 1), torch.bool)
# Topologically Sorted Source Nodes: [a_1], Original ATen: [aten.relu, aten.threshold_backward]
triton_poi_fused_relu_threshold_backward_0.run(buf3, primals_5, buf15, 1024, grid=grid(1024), stream=stream0)
del primals_5
buf4 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [linear_2], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_7, reinterpret_tensor(buf3, (64, 16), (16, 1), 0), reinterpret_tensor(primals_6, (16, 4), (1, 16), 0), alpha=1, beta=1, out=buf4)
del primals_7
buf5 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [linear_3], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_9, reinterpret_tensor(buf3, (64, 16), (16, 1), 0), reinterpret_tensor(primals_8, (16, 4), (1, 16), 0), alpha=1, beta=1, out=buf5)
del primals_9
buf6 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [tanh, mu], Original ATen: [aten.tanh, aten.mul]
triton_poi_fused_mul_tanh_1.run(buf4, buf6, 256, grid=grid(256), stream=stream0)
buf7 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
buf10 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [sigma, log, action_entropy], Original ATen: [aten.softplus, aten.log, aten.add]
triton_poi_fused_add_log_softplus_2.run(buf5, buf7, buf10, 256, grid=grid(256), stream=stream0)
# Topologically Sorted Source Nodes: [tanh, mu, sigma, action], Original ATen: [aten.tanh, aten.mul, aten.softplus, aten.normal]
buf8 = torch.ops.aten.normal.Tensor_Tensor(buf6, buf7)
buf9 = buf8
del buf8
buf11 = buf9; del buf9 # reuse
buf12 = buf6; del buf6 # reuse
buf13 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
buf14 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [log, action_1, var, sub, pow_2, neg, mul_1, truediv, sub_1, action_log_prob], Original ATen: [aten.log, aten.clamp, aten.pow, aten.sub, aten.neg, aten.mul, aten.div]
triton_poi_fused_clamp_div_log_mul_neg_pow_sub_3.run(buf11, buf12, buf7, buf13, buf14, 256, grid=grid(256), stream=stream0)
del buf7
return (buf11, buf14, buf10, reinterpret_tensor(primals_3, (64, 4), (4, 1), 0), reinterpret_tensor(buf1, (64, 16), (16, 1), 0), reinterpret_tensor(buf3, (64, 16), (16, 1), 0), buf4, buf5, buf12, buf13, primals_8, primals_6, buf15, primals_4, buf16, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((16, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((16, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((16, 16), (16, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((16, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((4, 16), (16, 1), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_8 = rand_strided((4, 16), (16, 1), device='cuda:0', dtype=torch.float32)
primals_9 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch as t
import torch.nn as nn
from torch.distributions import Normal
import torch.nn.functional as F
class A2CActorCont(nn.Module):
def __init__(self, state_dim, action_dim, action_range):
super().__init__()
self.fc1 = nn.Linear(state_dim, 16)
self.fc2 = nn.Linear(16, 16)
self.mu_head = nn.Linear(16, action_dim)
self.sigma_head = nn.Linear(16, action_dim)
self.action_range = action_range
def forward(self, state, action=None):
a = t.relu(self.fc1(state))
a = t.relu(self.fc2(a))
mu = 2.0 * t.tanh(self.mu_head(a))
sigma = F.softplus(self.sigma_head(a))
dist = Normal(mu, sigma)
action = action if action is not None else dist.sample()
action_entropy = dist.entropy()
action = action.clamp(-self.action_range, self.action_range)
action_log_prob = dist.log_prob(action)
return action, action_log_prob, action_entropy
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'state_dim': 4, 'action_dim': 4, 'action_range': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_relu_threshold_backward_0(in_out_ptr0, in_ptr0,
out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 1024
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 16
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(in_out_ptr0 + x2, tmp4, xmask)
tl.store(out_ptr0 + x2, tmp6, xmask)
@triton.jit
def triton_poi_fused_mul_tanh_1(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + x0, xmask)
tmp1 = libdevice.tanh(tmp0)
tmp2 = 2.0
tmp3 = tmp1 * tmp2
tl.store(out_ptr0 + x0, tmp3, xmask)
@triton.jit
def triton_poi_fused_add_log_softplus_2(in_ptr0, out_ptr0, out_ptr1, xnumel,
XBLOCK: tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + x0, xmask)
tmp1 = 20.0
tmp2 = tmp0 > tmp1
tmp3 = tl_math.exp(tmp0)
tmp4 = libdevice.log1p(tmp3)
tmp5 = tl.where(tmp2, tmp0, tmp4)
tmp6 = tl_math.log(tmp5)
tmp7 = 1.4189385332046727
tmp8 = tmp6 + tmp7
tl.store(out_ptr0 + x0, tmp5, xmask)
tl.store(out_ptr1 + x0, tmp8, xmask)
@triton.jit
def triton_poi_fused_clamp_div_log_mul_neg_pow_sub_3(in_out_ptr0,
in_out_ptr1, in_ptr0, out_ptr0, out_ptr1, xnumel, XBLOCK: tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_out_ptr0 + x0, xmask)
tmp5 = tl.load(in_out_ptr1 + x0, xmask)
tmp7 = tl.load(in_ptr0 + x0, xmask)
tmp1 = -4.0
tmp2 = triton_helpers.maximum(tmp0, tmp1)
tmp3 = 4.0
tmp4 = triton_helpers.minimum(tmp2, tmp3)
tmp6 = tmp4 - tmp5
tmp8 = tmp7 * tmp7
tmp9 = 2.0
tmp10 = tmp8 * tmp9
tmp11 = tmp6 * tmp6
tmp12 = -tmp11
tmp13 = tmp12 / tmp10
tmp14 = tl_math.log(tmp7)
tmp15 = tmp13 - tmp14
tmp16 = 0.9189385332046727
tmp17 = tmp15 - tmp16
tl.store(in_out_ptr0 + x0, tmp4, xmask)
tl.store(in_out_ptr1 + x0, tmp6, xmask)
tl.store(out_ptr0 + x0, tmp10, xmask)
tl.store(out_ptr1 + x0, tmp17, xmask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7, primals_8, primals_9) = args
args.clear()
assert_size_stride(primals_1, (16, 4), (4, 1))
assert_size_stride(primals_2, (16,), (1,))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_4, (16, 16), (16, 1))
assert_size_stride(primals_5, (16,), (1,))
assert_size_stride(primals_6, (4, 16), (16, 1))
assert_size_stride(primals_7, (4,), (1,))
assert_size_stride(primals_8, (4, 16), (16, 1))
assert_size_stride(primals_9, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 16), (16, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_3, (64, 4), (4, 1), 0),
reinterpret_tensor(primals_1, (4, 16), (1, 4), 0), out=buf0)
del primals_1
buf1 = reinterpret_tensor(buf0, (4, 4, 4, 16), (256, 64, 16, 1), 0)
del buf0
buf16 = empty_strided_cuda((4, 4, 4, 16), (256, 64, 16, 1), torch.bool)
get_raw_stream(0)
triton_poi_fused_relu_threshold_backward_0[grid(1024)](buf1,
primals_2, buf16, 1024, XBLOCK=128, num_warps=4, num_stages=1)
del primals_2
buf2 = empty_strided_cuda((64, 16), (16, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(buf1, (64, 16), (16, 1), 0),
reinterpret_tensor(primals_4, (16, 16), (1, 16), 0), out=buf2)
buf3 = reinterpret_tensor(buf2, (4, 4, 4, 16), (256, 64, 16, 1), 0)
del buf2
buf15 = empty_strided_cuda((4, 4, 4, 16), (256, 64, 16, 1), torch.bool)
triton_poi_fused_relu_threshold_backward_0[grid(1024)](buf3,
primals_5, buf15, 1024, XBLOCK=128, num_warps=4, num_stages=1)
del primals_5
buf4 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
extern_kernels.addmm(primals_7, reinterpret_tensor(buf3, (64, 16),
(16, 1), 0), reinterpret_tensor(primals_6, (16, 4), (1, 16), 0),
alpha=1, beta=1, out=buf4)
del primals_7
buf5 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
extern_kernels.addmm(primals_9, reinterpret_tensor(buf3, (64, 16),
(16, 1), 0), reinterpret_tensor(primals_8, (16, 4), (1, 16), 0),
alpha=1, beta=1, out=buf5)
del primals_9
buf6 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
triton_poi_fused_mul_tanh_1[grid(256)](buf4, buf6, 256, XBLOCK=128,
num_warps=4, num_stages=1)
buf7 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
buf10 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
triton_poi_fused_add_log_softplus_2[grid(256)](buf5, buf7, buf10,
256, XBLOCK=256, num_warps=4, num_stages=1)
buf8 = torch.ops.aten.normal.Tensor_Tensor(buf6, buf7)
buf9 = buf8
del buf8
buf11 = buf9
del buf9
buf12 = buf6
del buf6
buf13 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
buf14 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
triton_poi_fused_clamp_div_log_mul_neg_pow_sub_3[grid(256)](buf11,
buf12, buf7, buf13, buf14, 256, XBLOCK=128, num_warps=4,
num_stages=1)
del buf7
return buf11, buf14, buf10, reinterpret_tensor(primals_3, (64, 4), (4,
1), 0), reinterpret_tensor(buf1, (64, 16), (16, 1), 0
), reinterpret_tensor(buf3, (64, 16), (16, 1), 0
), buf4, buf5, buf12, buf13, primals_8, primals_6, buf15, primals_4, buf16
class A2CActorContNew(nn.Module):
def __init__(self, state_dim, action_dim, action_range):
super().__init__()
self.fc1 = nn.Linear(state_dim, 16)
self.fc2 = nn.Linear(16, 16)
self.mu_head = nn.Linear(16, action_dim)
self.sigma_head = nn.Linear(16, action_dim)
self.action_range = action_range
def forward(self, input_0):
primals_1 = self.fc1.weight
primals_2 = self.fc1.bias
primals_4 = self.fc2.weight
primals_5 = self.fc2.bias
primals_6 = self.mu_head.weight
primals_7 = self.mu_head.bias
primals_8 = self.sigma_head.weight
primals_9 = self.sigma_head.bias
primals_3 = input_0
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7, primals_8, primals_9])
return output[0], output[1], output[2]
| ikamensh/machin | A2CActorCont | false | 6,865 | [
"MIT"
] | 1 | af7b423c47bc1412530cf6c96c11bd3af9b3e239 | https://github.com/ikamensh/machin/tree/af7b423c47bc1412530cf6c96c11bd3af9b3e239 | import torch
import torch as t
import torch.nn as nn
from torch.distributions import Normal
import torch.nn.functional as F
class Model(nn.Module):
def __init__(self, state_dim, action_dim, action_range):
super().__init__()
self.fc1 = nn.Linear(state_dim, 16)
self.fc2 = nn.Linear(16, 16)
self.mu_head = nn.Linear(16, action_dim)
self.sigma_head = nn.Linear(16, action_dim)
self.action_range = action_range
def forward(self, state, action=None):
a = t.relu(self.fc1(state))
a = t.relu(self.fc2(a))
mu = 2.0 * t.tanh(self.mu_head(a))
sigma = F.softplus(self.sigma_head(a))
dist = Normal(mu, sigma)
action = action if action is not None else dist.sample()
action_entropy = dist.entropy()
action = action.clamp(-self.action_range, self.action_range)
action_log_prob = dist.log_prob(action)
return action, action_log_prob, action_entropy
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4, 4, 4]
|
Attention | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/av/cavn4qgxnuj4udauolvrrztemvv5rwe556xnoznnhqrxniumzyl3.py
# Topologically Sorted Source Nodes: [w_1], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# w_1 => exp
# Graph fragment:
# %mul_tensor : [num_users=2] = call_function[target=torch.ops.aten.mul.Tensor](args = (%bmm, 1), kwargs = {})
# %amax_default : [num_users=1] = call_function[target=torch.ops.aten.amax.default](args = (%mul_tensor, [-1], True), kwargs = {})
# %sub_tensor : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%mul_tensor, %amax_default), kwargs = {})
# %div_tensor : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%sub_tensor, 8.0), kwargs = {})
# %exp : [num_users=2] = call_function[target=torch.ops.aten.exp.default](args = (%div_tensor,), kwargs = {})
triton_poi_fused__softmax_0 = async_compile.triton('triton_poi_fused__softmax_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_0(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp3 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp8 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp11 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp1 = 1.0
tmp2 = tmp0 * tmp1
tmp4 = tmp3 * tmp1
tmp6 = tmp5 * tmp1
tmp7 = triton_helpers.maximum(tmp4, tmp6)
tmp9 = tmp8 * tmp1
tmp10 = triton_helpers.maximum(tmp7, tmp9)
tmp12 = tmp11 * tmp1
tmp13 = triton_helpers.maximum(tmp10, tmp12)
tmp14 = tmp2 - tmp13
tmp15 = 0.125
tmp16 = tmp14 * tmp15
tmp17 = tl_math.exp(tmp16)
tl.store(out_ptr0 + (x2), tmp17, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/kj/ckjtlefzavjukjsytvkak6ek26zmzexpcbnlwelx4k5kascjxlf3.py
# Topologically Sorted Source Nodes: [w_1], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# w_1 => div_1, sum_1
# Graph fragment:
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%exp, [-1], True), kwargs = {})
# %div_1 : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%exp, %sum_1), kwargs = {})
triton_poi_fused__softmax_1 = async_compile.triton('triton_poi_fused__softmax_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_1(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + (x2), tmp8, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, arg1_1, arg2_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4), (16, 4, 1))
assert_size_stride(arg1_1, (4, 4, 4), (16, 4, 1))
assert_size_stride(arg2_1, (4, 4, 4), (16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [w], Original ATen: [aten.bmm]
extern_kernels.bmm(arg1_1, reinterpret_tensor(arg0_1, (4, 4, 4), (16, 1, 4), 0), out=buf0)
del arg0_1
del arg1_1
buf1 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [w_1], Original ATen: [aten._softmax]
stream0 = get_raw_stream(0)
triton_poi_fused__softmax_0.run(buf0, buf1, 64, grid=grid(64), stream=stream0)
buf2 = buf0; del buf0 # reuse
# Topologically Sorted Source Nodes: [w_1], Original ATen: [aten._softmax]
triton_poi_fused__softmax_1.run(buf1, buf2, 64, grid=grid(64), stream=stream0)
buf3 = buf1; del buf1 # reuse
# Topologically Sorted Source Nodes: [w_1, c], Original ATen: [aten._softmax, aten.bmm]
extern_kernels.bmm(buf2, arg2_1, out=buf3)
del arg2_1
del buf2
return (buf3, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
arg1_1 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
arg2_1 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([arg0_1, arg1_1, arg2_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
class Attention(nn.Module):
def __init__(self):
super().__init__()
self.softmax = nn.Softmax(dim=-1)
def forward(self, Q, K, V, mask=None, dk=64):
w = torch.bmm(Q, K.transpose(1, 2))
if mask is not None:
assert w.size() == mask.size()
w.masked_fill_(mask, -float('inf'))
w = self.softmax(w / dk ** 0.5)
c = torch.bmm(w, V)
return c
def get_inputs():
return [torch.rand([4, 4, 4]), torch.rand([4, 4, 4]), torch.rand([4, 4, 4])
]
def get_init_inputs():
return [[], {}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import math as tl_math
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused__softmax_0(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp3 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp8 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp11 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last'
)
tmp1 = 1.0
tmp2 = tmp0 * tmp1
tmp4 = tmp3 * tmp1
tmp6 = tmp5 * tmp1
tmp7 = triton_helpers.maximum(tmp4, tmp6)
tmp9 = tmp8 * tmp1
tmp10 = triton_helpers.maximum(tmp7, tmp9)
tmp12 = tmp11 * tmp1
tmp13 = triton_helpers.maximum(tmp10, tmp12)
tmp14 = tmp2 - tmp13
tmp15 = 0.125
tmp16 = tmp14 * tmp15
tmp17 = tl_math.exp(tmp16)
tl.store(out_ptr0 + x2, tmp17, xmask)
@triton.jit
def triton_poi_fused__softmax_1(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + x2, tmp8, xmask)
def call(args):
arg0_1, arg1_1, arg2_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4), (16, 4, 1))
assert_size_stride(arg1_1, (4, 4, 4), (16, 4, 1))
assert_size_stride(arg2_1, (4, 4, 4), (16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
extern_kernels.bmm(arg1_1, reinterpret_tensor(arg0_1, (4, 4, 4), (
16, 1, 4), 0), out=buf0)
del arg0_1
del arg1_1
buf1 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused__softmax_0[grid(64)](buf0, buf1, 64, XBLOCK=64,
num_warps=1, num_stages=1)
buf2 = buf0
del buf0
triton_poi_fused__softmax_1[grid(64)](buf1, buf2, 64, XBLOCK=64,
num_warps=1, num_stages=1)
buf3 = buf1
del buf1
extern_kernels.bmm(buf2, arg2_1, out=buf3)
del arg2_1
del buf2
return buf3,
class AttentionNew(nn.Module):
def __init__(self):
super().__init__()
self.softmax = nn.Softmax(dim=-1)
def forward(self, input_0, input_1, input_2):
arg0_1 = input_0
arg1_1 = input_1
arg2_1 = input_2
output = call([arg0_1, arg1_1, arg2_1])
return output[0]
| iml1111/machine-translation | Attention | false | 6,866 | [
"MIT"
] | 1 | a7dd673efbe8a172c1df49e0d50482dc84008c37 | https://github.com/iml1111/machine-translation/tree/a7dd673efbe8a172c1df49e0d50482dc84008c37 | import torch
import torch.nn as nn
class Model(nn.Module):
def __init__(self):
super().__init__()
self.softmax = nn.Softmax(dim=-1)
def forward(self, Q, K, V, mask=None, dk=64):
w = torch.bmm(Q, K.transpose(1, 2))
if mask is not None:
assert w.size() == mask.size()
w.masked_fill_(mask, -float('inf'))
w = self.softmax(w / dk ** 0.5)
c = torch.bmm(w, V)
return c
def get_inputs():
return [torch.rand([4, 4, 4]), torch.rand([4, 4, 4]), torch.rand([4, 4, 4])
]
def get_init_inputs():
return []
|
HSwishV2 | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/y6/cy64rbflt45qv6aforjcbtkgydz4w7xgtikbd34mvs7m4l5k2beu.py
# Topologically Sorted Source Nodes: [add, relu6, mul_, act], Original ATen: [aten.add, aten.hardtanh, aten.mul, aten.div]
# Source node to ATen node mapping:
# act => div
# add => add
# mul_ => mul
# relu6 => clamp_max, clamp_min
# Graph fragment:
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%arg0_1, 3), kwargs = {})
# %clamp_min : [num_users=1] = call_function[target=torch.ops.aten.clamp_min.default](args = (%add, 0), kwargs = {})
# %clamp_max : [num_users=1] = call_function[target=torch.ops.aten.clamp_max.default](args = (%clamp_min, 6), kwargs = {})
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%clamp_max, %arg0_1), kwargs = {})
# %div : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%mul, 6), kwargs = {})
triton_poi_fused_add_div_hardtanh_mul_0 = async_compile.triton('triton_poi_fused_add_div_hardtanh_mul_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_div_hardtanh_mul_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_div_hardtanh_mul_0(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (x0), xmask)
tmp1 = 3.0
tmp2 = tmp0 + tmp1
tmp3 = 0.0
tmp4 = triton_helpers.maximum(tmp2, tmp3)
tmp5 = 6.0
tmp6 = triton_helpers.minimum(tmp4, tmp5)
tmp7 = tmp6 * tmp0
tmp8 = 0.16666666666666666
tmp9 = tmp7 * tmp8
tl.store(out_ptr0 + (x0), tmp9, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [add, relu6, mul_, act], Original ATen: [aten.add, aten.hardtanh, aten.mul, aten.div]
stream0 = get_raw_stream(0)
triton_poi_fused_add_div_hardtanh_mul_0.run(arg0_1, buf0, 256, grid=grid(256), stream=stream0)
del arg0_1
return (buf0, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([arg0_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.nn.functional as F
class HSwishFunctionV2(torch.autograd.Function):
@staticmethod
def forward(ctx, feat):
act = F.relu6(feat + 3).mul_(feat).div_(6)
ctx.variables = feat
return act
@staticmethod
def backward(ctx, grad_output):
feat = ctx.variables
grad = F.relu6(feat + 3).div_(6)
grad.add_(torch.where(torch.eq(-3 < feat, feat < 3), torch.
ones_like(feat).div_(6), torch.zeros_like(feat)).mul_(feat))
grad *= grad_output
return grad
class HSwishV2(nn.Module):
def __init__(self):
super(HSwishV2, self).__init__()
def forward(self, feat):
return HSwishFunctionV2.apply(feat)
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {}]
| import torch
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
import torch.nn as nn
import torch.nn.functional as F
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_poi_fused_add_div_hardtanh_mul_0(in_ptr0, out_ptr0, xnumel,
XBLOCK: tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + x0, xmask)
tmp1 = 3.0
tmp2 = tmp0 + tmp1
tmp3 = 0.0
tmp4 = triton_helpers.maximum(tmp2, tmp3)
tmp5 = 6.0
tmp6 = triton_helpers.minimum(tmp4, tmp5)
tmp7 = tmp6 * tmp0
tmp8 = 0.16666666666666666
tmp9 = tmp7 * tmp8
tl.store(out_ptr0 + x0, tmp9, xmask)
def call(args):
arg0_1, = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_add_div_hardtanh_mul_0[grid(256)](arg0_1, buf0,
256, XBLOCK=128, num_warps=4, num_stages=1)
del arg0_1
return buf0,
class HSwishFunctionV2(torch.autograd.Function):
@staticmethod
def forward(ctx, feat):
act = F.relu6(feat + 3).mul_(feat).div_(6)
ctx.variables = feat
return act
@staticmethod
def backward(ctx, grad_output):
feat = ctx.variables
grad = F.relu6(feat + 3).div_(6)
grad.add_(torch.where(torch.eq(-3 < feat, feat < 3), torch.
ones_like(feat).div_(6), torch.zeros_like(feat)).mul_(feat))
grad *= grad_output
return grad
class HSwishV2New(nn.Module):
def __init__(self):
super(HSwishV2New, self).__init__()
def forward(self, input_0):
arg0_1 = input_0
output = call([arg0_1])
return output[0]
| imvladikon/pytorch-loss | HSwishV2 | false | 6,867 | [
"MIT"
] | 1 | 6cfaabe1be898e1ff000b3dffb46d0ef09096f6b | https://github.com/imvladikon/pytorch-loss/tree/6cfaabe1be898e1ff000b3dffb46d0ef09096f6b | import torch
import torch.nn as nn
import torch.nn.functional as F
class HSwishFunctionV2(torch.autograd.Function):
@staticmethod
def forward(ctx, feat):
act = F.relu6(feat + 3).mul_(feat).div_(6)
ctx.variables = feat
return act
@staticmethod
def backward(ctx, grad_output):
feat = ctx.variables
grad = F.relu6(feat + 3).div_(6)
grad.add_(torch.where(torch.eq(-3 < feat, feat < 3), torch.
ones_like(feat).div_(6), torch.zeros_like(feat)).mul_(feat))
grad *= grad_output
return grad
class Model(nn.Module):
def __init__(self):
super().__init__()
def forward(self, feat):
return HSwishFunctionV2.apply(feat)
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return []
|
TemperatureHolder | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/xp/cxpd2bcgjjguvsbnowhprcscim22tcfdskii4vimpwk2r73oeben.py
# Topologically Sorted Source Nodes: [exp], Original ATen: [aten.exp]
# Source node to ATen node mapping:
# exp => exp
# Graph fragment:
# %exp : [num_users=1] = call_function[target=torch.ops.aten.exp.default](args = (%primals_1,), kwargs = {})
triton_poi_fused_exp_0 = async_compile.triton('triton_poi_fused_exp_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[1],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {2: 1}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1), equal_to_1=(2,))]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_exp_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_exp_0(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 1
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = tl.full([XBLOCK], True, tl.int1)
tmp0 = tl.load(in_ptr0 + (0))
tmp1 = tl.broadcast_to(tmp0, [XBLOCK])
tmp2 = tl_math.exp(tmp1)
tl.store(out_ptr0 + (tl.full([XBLOCK], 0, tl.int32)), tmp2, None)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, = args
args.clear()
assert_size_stride(primals_1, (), ())
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((), (), torch.float32)
# Topologically Sorted Source Nodes: [exp], Original ATen: [aten.exp]
stream0 = get_raw_stream(0)
triton_poi_fused_exp_0.run(primals_1, buf0, 1, grid=grid(1), stream=stream0)
del primals_1
return (buf0, buf0, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((), (), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
class TemperatureHolder(nn.Module):
"""Module that holds a temperature as a learnable value.
Args:
initial_log_temperature (float): Initial value of log(temperature).
"""
def __init__(self, initial_log_temperature=0):
super().__init__()
self.log_temperature = nn.Parameter(torch.tensor(
initial_log_temperature, dtype=torch.float32))
def forward(self):
"""Return a temperature as a torch.Tensor."""
return torch.exp(self.log_temperature)
def get_inputs():
return []
def get_init_inputs():
return [[], {}]
| import torch
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime.triton_helpers import math as tl_math
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_poi_fused_exp_0(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xoffset = tl.program_id(0) * XBLOCK
xoffset + tl.arange(0, XBLOCK)[:]
tl.full([XBLOCK], True, tl.int1)
tmp0 = tl.load(in_ptr0 + 0)
tmp1 = tl.broadcast_to(tmp0, [XBLOCK])
tmp2 = tl_math.exp(tmp1)
tl.store(out_ptr0 + tl.full([XBLOCK], 0, tl.int32), tmp2, None)
def call(args):
primals_1, = args
args.clear()
assert_size_stride(primals_1, (), ())
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((), (), torch.float32)
get_raw_stream(0)
triton_poi_fused_exp_0[grid(1)](primals_1, buf0, 1, XBLOCK=1,
num_warps=1, num_stages=1)
del primals_1
return buf0, buf0
class TemperatureHolderNew(nn.Module):
"""Module that holds a temperature as a learnable value.
Args:
initial_log_temperature (float): Initial value of log(temperature).
"""
def __init__(self, initial_log_temperature=0):
super().__init__()
self.log_temperature = nn.Parameter(torch.tensor(
initial_log_temperature, dtype=torch.float32))
def forward(self):
primals_1 = self.log_temperature
output = call([primals_1])
return output[0]
| imatge-upc/pixelcoordEDL | TemperatureHolder | false | 6,868 | [
"MIT"
] | 1 | 353632feed6ac8c93758c1a2a1b7a477e7ff053c | https://github.com/imatge-upc/pixelcoordEDL/tree/353632feed6ac8c93758c1a2a1b7a477e7ff053c | import torch
import torch.nn as nn
class Model(nn.Module):
"""Module that holds a temperature as a learnable value.
Args:
initial_log_temperature (float): Initial value of log(temperature).
"""
def __init__(self, initial_log_temperature=0):
super().__init__()
self.log_temperature = nn.Parameter(torch.tensor(
initial_log_temperature, dtype=torch.float32))
def forward(self):
"""Return a temperature as a torch.Tensor."""
return torch.exp(self.log_temperature)
def get_inputs():
return []
def get_init_inputs():
return []
|
AsymmetricLossOptimized | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/ry/cry5kqibd2j5v7zh6b5qbaeadqz6hqmokj5ujzqm6gt55y7roh7a.py
# Topologically Sorted Source Nodes: [sigmoid, clamp, log, mul, sub, sub_1, add_, clamp_, clamp_1, log_1, mul_1, add__1, mul_2, sub_2, mul_3, sub_3, mul_4, mul_5, add, pow_1, imul, sum_1, neg], Original ATen: [aten.sigmoid, aten.clamp, aten.log, aten.mul, aten.rsub, aten.add, aten.sub, aten.pow, aten.sum, aten.neg]
# Source node to ATen node mapping:
# add => add_2
# add_ => add
# add__1 => add_1
# clamp => clamp_min
# clamp_ => clamp_max
# clamp_1 => clamp_min_1
# imul => mul_6
# log => log
# log_1 => log_1
# mul => mul
# mul_1 => mul_1
# mul_2 => mul_2
# mul_3 => mul_3
# mul_4 => mul_4
# mul_5 => mul_5
# neg => neg
# pow_1 => pow_1
# sigmoid => sigmoid
# sub => sub
# sub_1 => sub_1
# sub_2 => sub_2
# sub_3 => sub_3
# sum_1 => sum_1
# Graph fragment:
# %sigmoid : [num_users=3] = call_function[target=torch.ops.aten.sigmoid.default](args = (%arg1_1,), kwargs = {})
# %clamp_min : [num_users=1] = call_function[target=torch.ops.aten.clamp_min.default](args = (%sigmoid, 1e-08), kwargs = {})
# %log : [num_users=1] = call_function[target=torch.ops.aten.log.default](args = (%clamp_min,), kwargs = {})
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%arg0_1, %log), kwargs = {})
# %sub : [num_users=4] = call_function[target=torch.ops.aten.sub.Tensor](args = (1, %arg0_1), kwargs = {})
# %sub_1 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (1.0, %sigmoid), kwargs = {})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%sub_1, 0.05), kwargs = {})
# %clamp_max : [num_users=2] = call_function[target=torch.ops.aten.clamp_max.default](args = (%add, 1), kwargs = {})
# %clamp_min_1 : [num_users=1] = call_function[target=torch.ops.aten.clamp_min.default](args = (%clamp_max, 1e-08), kwargs = {})
# %log_1 : [num_users=1] = call_function[target=torch.ops.aten.log.default](args = (%clamp_min_1,), kwargs = {})
# %mul_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sub, %log_1), kwargs = {})
# %add_1 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul, %mul_1), kwargs = {})
# %mul_2 : [num_users=2] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sigmoid, %arg0_1), kwargs = {})
# %sub_2 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (1, %mul_2), kwargs = {})
# %mul_3 : [num_users=2] = call_function[target=torch.ops.aten.mul.Tensor](args = (%clamp_max, %sub), kwargs = {})
# %sub_3 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%sub_2, %mul_3), kwargs = {})
# %mul_4 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%arg0_1, 1), kwargs = {})
# %mul_5 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sub, 4), kwargs = {})
# %add_2 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul_4, %mul_5), kwargs = {})
# %pow_1 : [num_users=2] = call_function[target=torch.ops.aten.pow.Tensor_Tensor](args = (%sub_3, %add_2), kwargs = {})
# %mul_6 : [num_users=2] = call_function[target=torch.ops.aten.mul.Tensor](args = (%add_1, %pow_1), kwargs = {})
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.default](args = (%mul_6,), kwargs = {})
# %neg : [num_users=1] = call_function[target=torch.ops.aten.neg.default](args = (%sum_1,), kwargs = {})
triton_per_fused_add_clamp_log_mul_neg_pow_rsub_sigmoid_sub_sum_0 = async_compile.triton('triton_per_fused_add_clamp_log_mul_neg_pow_rsub_sigmoid_sub_sum_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[1, 256],
reduction_hint=ReductionHint.INNER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: '*fp32', 5: '*fp32', 6: '*fp32', 7: '*fp32', 8: 'i32', 9: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {8: 1}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5, 6, 7, 9), equal_to_1=(8,))]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused_add_clamp_log_mul_neg_pow_rsub_sigmoid_sub_sum_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': True, 'num_load': 2, 'num_reduction': 1, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused_add_clamp_log_mul_neg_pow_rsub_sigmoid_sub_sum_0(in_out_ptr0, in_ptr0, in_ptr1, out_ptr0, out_ptr1, out_ptr2, out_ptr3, out_ptr4, xnumel, rnumel):
xnumel = 1
XBLOCK: tl.constexpr = 1
rnumel = 256
RBLOCK: tl.constexpr = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = tl.full([1], xoffset, tl.int32)
xmask = tl.full([RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[:]
roffset = 0
rmask = tl.full([RBLOCK], True, tl.int1)
r0 = rindex
tmp0 = tl.load(in_ptr0 + (r0), None)
tmp3 = tl.load(in_ptr1 + (r0), None)
tmp1 = 1.0
tmp2 = tmp1 - tmp0
tmp4 = tl.sigmoid(tmp3)
tmp5 = tmp4 * tmp0
tmp6 = tmp1 - tmp4
tmp7 = 0.05
tmp8 = tmp6 + tmp7
tmp9 = triton_helpers.minimum(tmp8, tmp1)
tmp10 = tmp9 * tmp2
tmp11 = tmp1 - tmp5
tmp12 = tmp11 - tmp10
tmp13 = tmp0 * tmp1
tmp14 = 4.0
tmp15 = tmp2 * tmp14
tmp16 = tmp13 + tmp15
tmp17 = libdevice.pow(tmp12, tmp16)
tmp18 = 1e-08
tmp19 = triton_helpers.maximum(tmp4, tmp18)
tmp20 = tl_math.log(tmp19)
tmp21 = tmp0 * tmp20
tmp22 = triton_helpers.maximum(tmp9, tmp18)
tmp23 = tl_math.log(tmp22)
tmp24 = tmp2 * tmp23
tmp25 = tmp21 + tmp24
tmp26 = tmp25 * tmp17
tmp27 = tl.broadcast_to(tmp26, [RBLOCK])
tmp29 = triton_helpers.promote_to_tensor(tl.sum(tmp27, 0))
tmp30 = -tmp29
tl.store(out_ptr0 + (tl.broadcast_to(r0, [RBLOCK])), tmp2, None)
tl.store(out_ptr1 + (tl.broadcast_to(r0, [RBLOCK])), tmp5, None)
tl.store(out_ptr2 + (tl.broadcast_to(r0, [RBLOCK])), tmp10, None)
tl.store(out_ptr3 + (tl.broadcast_to(r0, [RBLOCK])), tmp17, None)
tl.store(out_ptr4 + (tl.broadcast_to(r0, [RBLOCK])), tmp26, None)
tl.debug_barrier()
tl.store(in_out_ptr0 + (tl.full([1], 0, tl.int32)), tmp30, None)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, arg1_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg1_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
buf1 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
buf2 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
buf3 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
buf4 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
buf5 = empty_strided_cuda((), (), torch.float32)
buf6 = buf5; del buf5 # reuse
# Topologically Sorted Source Nodes: [sigmoid, clamp, log, mul, sub, sub_1, add_, clamp_, clamp_1, log_1, mul_1, add__1, mul_2, sub_2, mul_3, sub_3, mul_4, mul_5, add, pow_1, imul, sum_1, neg], Original ATen: [aten.sigmoid, aten.clamp, aten.log, aten.mul, aten.rsub, aten.add, aten.sub, aten.pow, aten.sum, aten.neg]
stream0 = get_raw_stream(0)
triton_per_fused_add_clamp_log_mul_neg_pow_rsub_sigmoid_sub_sum_0.run(buf6, arg0_1, arg1_1, buf0, buf1, buf2, buf3, buf4, 1, 256, grid=grid(1), stream=stream0)
del arg0_1
del arg1_1
return (buf6, buf3, buf4, buf2, buf1, buf0, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
arg1_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([arg0_1, arg1_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
class AsymmetricLossOptimized(nn.Module):
""" Notice - optimized version, minimizes memory allocation and gpu uploading,
favors inplace operations"""
def __init__(self, gamma_neg=4, gamma_pos=1, clip=0.05, eps=1e-08,
disable_torch_grad_focal_loss=False):
super(AsymmetricLossOptimized, self).__init__()
self.gamma_neg = gamma_neg
self.gamma_pos = gamma_pos
self.clip = clip
self.disable_torch_grad_focal_loss = disable_torch_grad_focal_loss
self.eps = eps
(self.targets) = (self.anti_targets) = (self.xs_pos) = (self.xs_neg
) = (self.asymmetric_w) = (self.loss) = None
def forward(self, x, y):
""""
Parameters
----------
x: input logits
y: targets (multi-label binarized vector)
"""
self.targets = y
self.anti_targets = 1 - y
self.xs_pos = torch.sigmoid(x)
self.xs_neg = 1.0 - self.xs_pos
if self.clip is not None and self.clip > 0:
self.xs_neg.add_(self.clip).clamp_(max=1)
self.loss = self.targets * torch.log(self.xs_pos.clamp(min=self.eps))
self.loss.add_(self.anti_targets * torch.log(self.xs_neg.clamp(min=
self.eps)))
if self.gamma_neg > 0 or self.gamma_pos > 0:
if self.disable_torch_grad_focal_loss:
torch._C.set_grad_enabled(False)
self.xs_pos = self.xs_pos * self.targets
self.xs_neg = self.xs_neg * self.anti_targets
self.asymmetric_w = torch.pow(1 - self.xs_pos - self.xs_neg,
self.gamma_pos * self.targets + self.gamma_neg * self.
anti_targets)
if self.disable_torch_grad_focal_loss:
torch._C.set_grad_enabled(True)
self.loss *= self.asymmetric_w
return -self.loss.sum()
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {}]
| import torch
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_per_fused_add_clamp_log_mul_neg_pow_rsub_sigmoid_sub_sum_0(
in_out_ptr0, in_ptr0, in_ptr1, out_ptr0, out_ptr1, out_ptr2, out_ptr3,
out_ptr4, xnumel, rnumel):
XBLOCK: tl.constexpr = 1
RBLOCK: tl.constexpr = 256
xoffset = tl.program_id(0) * XBLOCK
tl.full([1], xoffset, tl.int32)
tl.full([RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[:]
tl.full([RBLOCK], True, tl.int1)
r0 = rindex
tmp0 = tl.load(in_ptr0 + r0, None)
tmp3 = tl.load(in_ptr1 + r0, None)
tmp1 = 1.0
tmp2 = tmp1 - tmp0
tmp4 = tl.sigmoid(tmp3)
tmp5 = tmp4 * tmp0
tmp6 = tmp1 - tmp4
tmp7 = 0.05
tmp8 = tmp6 + tmp7
tmp9 = triton_helpers.minimum(tmp8, tmp1)
tmp10 = tmp9 * tmp2
tmp11 = tmp1 - tmp5
tmp12 = tmp11 - tmp10
tmp13 = tmp0 * tmp1
tmp14 = 4.0
tmp15 = tmp2 * tmp14
tmp16 = tmp13 + tmp15
tmp17 = libdevice.pow(tmp12, tmp16)
tmp18 = 1e-08
tmp19 = triton_helpers.maximum(tmp4, tmp18)
tmp20 = tl_math.log(tmp19)
tmp21 = tmp0 * tmp20
tmp22 = triton_helpers.maximum(tmp9, tmp18)
tmp23 = tl_math.log(tmp22)
tmp24 = tmp2 * tmp23
tmp25 = tmp21 + tmp24
tmp26 = tmp25 * tmp17
tmp27 = tl.broadcast_to(tmp26, [RBLOCK])
tmp29 = triton_helpers.promote_to_tensor(tl.sum(tmp27, 0))
tmp30 = -tmp29
tl.store(out_ptr0 + tl.broadcast_to(r0, [RBLOCK]), tmp2, None)
tl.store(out_ptr1 + tl.broadcast_to(r0, [RBLOCK]), tmp5, None)
tl.store(out_ptr2 + tl.broadcast_to(r0, [RBLOCK]), tmp10, None)
tl.store(out_ptr3 + tl.broadcast_to(r0, [RBLOCK]), tmp17, None)
tl.store(out_ptr4 + tl.broadcast_to(r0, [RBLOCK]), tmp26, None)
tl.debug_barrier()
tl.store(in_out_ptr0 + tl.full([1], 0, tl.int32), tmp30, None)
def call(args):
arg0_1, arg1_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg1_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
buf1 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
buf2 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
buf3 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
buf4 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
buf5 = empty_strided_cuda((), (), torch.float32)
buf6 = buf5
del buf5
get_raw_stream(0)
triton_per_fused_add_clamp_log_mul_neg_pow_rsub_sigmoid_sub_sum_0[grid
(1)](buf6, arg0_1, arg1_1, buf0, buf1, buf2, buf3, buf4, 1, 256,
num_warps=2, num_stages=1)
del arg0_1
del arg1_1
return buf6, buf3, buf4, buf2, buf1, buf0
class AsymmetricLossOptimizedNew(nn.Module):
""" Notice - optimized version, minimizes memory allocation and gpu uploading,
favors inplace operations"""
def __init__(self, gamma_neg=4, gamma_pos=1, clip=0.05, eps=1e-08,
disable_torch_grad_focal_loss=False):
super(AsymmetricLossOptimizedNew, self).__init__()
self.gamma_neg = gamma_neg
self.gamma_pos = gamma_pos
self.clip = clip
self.disable_torch_grad_focal_loss = disable_torch_grad_focal_loss
self.eps = eps
(self.targets) = (self.anti_targets) = (self.xs_pos) = (self.xs_neg
) = (self.asymmetric_w) = (self.loss) = None
def forward(self, input_0, input_1):
arg0_1 = input_0
arg1_1 = input_1
output = call([arg0_1, arg1_1])
return output[0]
| imvladikon/pytorch-loss | AsymmetricLossOptimized | false | 6,869 | [
"MIT"
] | 1 | 6cfaabe1be898e1ff000b3dffb46d0ef09096f6b | https://github.com/imvladikon/pytorch-loss/tree/6cfaabe1be898e1ff000b3dffb46d0ef09096f6b | import torch
import torch.nn as nn
class Model(nn.Module):
""" Notice - optimized version, minimizes memory allocation and gpu uploading,
favors inplace operations"""
def __init__(self, gamma_neg=4, gamma_pos=1, clip=0.05, eps=1e-08,
disable_torch_grad_focal_loss=False):
super().__init__()
self.gamma_neg = gamma_neg
self.gamma_pos = gamma_pos
self.clip = clip
self.disable_torch_grad_focal_loss = disable_torch_grad_focal_loss
self.eps = eps
(self.targets) = (self.anti_targets) = (self.xs_pos) = (self.xs_neg
) = (self.asymmetric_w) = (self.loss) = None
def forward(self, x, y):
""""
Parameters
----------
x: input logits
y: targets (multi-label binarized vector)
"""
self.targets = y
self.anti_targets = 1 - y
self.xs_pos = torch.sigmoid(x)
self.xs_neg = 1.0 - self.xs_pos
if self.clip is not None and self.clip > 0:
self.xs_neg.add_(self.clip).clamp_(max=1)
self.loss = self.targets * torch.log(self.xs_pos.clamp(min=self.eps))
self.loss.add_(self.anti_targets * torch.log(self.xs_neg.clamp(min=
self.eps)))
if self.gamma_neg > 0 or self.gamma_pos > 0:
if self.disable_torch_grad_focal_loss:
torch._C.set_grad_enabled(False)
self.xs_pos = self.xs_pos * self.targets
self.xs_neg = self.xs_neg * self.anti_targets
self.asymmetric_w = torch.pow(1 - self.xs_pos - self.xs_neg,
self.gamma_pos * self.targets + self.gamma_neg * self.
anti_targets)
if self.disable_torch_grad_focal_loss:
torch._C.set_grad_enabled(True)
self.loss *= self.asymmetric_w
return -self.loss.sum()
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return []
|
Self_Attentive_Pooling | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/yd/cydbtjoq352gcolmflbvu2nqkda7xg7q5hnvltb47jsg5dbmubym.py
# Topologically Sorted Source Nodes: [linear], Original ATen: [aten.clone]
# Source node to ATen node mapping:
# linear => clone
# Graph fragment:
# %clone : [num_users=1] = call_function[target=torch.ops.aten.clone.default](args = (%permute,), kwargs = {memory_format: torch.contiguous_format})
triton_poi_fused_clone_0 = async_compile.triton('triton_poi_fused_clone_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16, 4], tile_hint=TileHint.SQUARE,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clone_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clone_0(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = (yindex // 4)
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + (4*x2) + (16*y1)), xmask & ymask, eviction_policy='evict_last')
tl.store(out_ptr0 + (x2 + (4*y3)), tmp0, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/2g/c2gw7362i2a6wsfdx2sxyywx4o6ronjg6goebvdn44w6gpjsxpbc.py
# Topologically Sorted Source Nodes: [linear, h], Original ATen: [aten.add, aten.tanh]
# Source node to ATen node mapping:
# h => tanh
# linear => add
# Graph fragment:
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%view_1, %primals_3), kwargs = {})
# %tanh : [num_users=2] = call_function[target=torch.ops.aten.tanh.default](args = (%add,), kwargs = {})
triton_poi_fused_add_tanh_1 = async_compile.triton('triton_poi_fused_add_tanh_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_tanh_1', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_tanh_1(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = libdevice.tanh(tmp2)
tl.store(in_out_ptr0 + (x2), tmp3, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/gd/cgdtd7uw2iemby2kfb22fx3vkhdbrpyx2y2l6nq45fmox3ad7stv.py
# Topologically Sorted Source Nodes: [softmax], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# softmax => amax, exp, sub
# Graph fragment:
# %amax : [num_users=1] = call_function[target=torch.ops.aten.amax.default](args = (%squeeze, [1], True), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%squeeze, %amax), kwargs = {})
# %exp : [num_users=2] = call_function[target=torch.ops.aten.exp.default](args = (%sub,), kwargs = {})
triton_poi_fused__softmax_2 = async_compile.triton('triton_poi_fused__softmax_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_2(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp3 = triton_helpers.maximum(tmp1, tmp2)
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp7 = triton_helpers.maximum(tmp5, tmp6)
tmp8 = tmp0 - tmp7
tmp9 = tl_math.exp(tmp8)
tl.store(out_ptr0 + (x2), tmp9, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/qs/cqsyda2m63ct5ijcfgcipyyfn273chi5d3kmpjuf5asa7h4wdpdv.py
# Topologically Sorted Source Nodes: [softmax], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# softmax => div, sum_1
# Graph fragment:
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%exp, [1], True), kwargs = {})
# %div : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%exp, %sum_1), kwargs = {})
triton_poi_fused__softmax_3 = async_compile.triton('triton_poi_fused__softmax_3', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_3', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_3(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + (x2), tmp8, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/jc/cjcndhrjkkewstr5xgidpsgvesqj5zoe5axasmayxpubnwdcuadd.py
# Topologically Sorted Source Nodes: [mul, x_1], Original ATen: [aten.mul, aten.sum]
# Source node to ATen node mapping:
# mul => mul
# x_1 => sum_2
# Graph fragment:
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%permute, %view_4), kwargs = {})
# %sum_2 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%mul, [1]), kwargs = {})
triton_poi_fused_mul_sum_4 = async_compile.triton('triton_poi_fused_mul_sum_4', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_mul_sum_4', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 8, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_mul_sum_4(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (4*x2), xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (4*x1), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr0 + (1 + (4*x2)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr1 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr0 + (2 + (4*x2)), xmask, eviction_policy='evict_last')
tmp8 = tl.load(in_ptr1 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp11 = tl.load(in_ptr0 + (3 + (4*x2)), xmask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr1 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp2 = tmp0 * tmp1
tmp5 = tmp3 * tmp4
tmp6 = tmp2 + tmp5
tmp9 = tmp7 * tmp8
tmp10 = tmp6 + tmp9
tmp13 = tmp11 * tmp12
tmp14 = tmp10 + tmp13
tl.store(out_ptr0 + (x2), tmp14, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_2, (4, 4), (4, 1))
assert_size_stride(primals_3, (4, ), (1, ))
assert_size_stride(primals_4, (4, 1), (1, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [linear], Original ATen: [aten.clone]
stream0 = get_raw_stream(0)
triton_poi_fused_clone_0.run(primals_1, buf0, 16, 4, grid=grid(16, 4), stream=stream0)
buf1 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [linear], Original ATen: [aten.mm]
extern_kernels.mm(reinterpret_tensor(buf0, (16, 4), (4, 1), 0), reinterpret_tensor(primals_2, (4, 4), (1, 4), 0), out=buf1)
del primals_2
buf2 = reinterpret_tensor(buf1, (4, 4, 4), (16, 4, 1), 0); del buf1 # reuse
# Topologically Sorted Source Nodes: [linear, h], Original ATen: [aten.add, aten.tanh]
triton_poi_fused_add_tanh_1.run(buf2, primals_3, 64, grid=grid(64), stream=stream0)
del primals_3
buf3 = empty_strided_cuda((16, 1), (1, 1), torch.float32)
# Topologically Sorted Source Nodes: [matmul], Original ATen: [aten.mm]
extern_kernels.mm(reinterpret_tensor(buf2, (16, 4), (4, 1), 0), primals_4, out=buf3)
buf4 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [softmax], Original ATen: [aten._softmax]
triton_poi_fused__softmax_2.run(buf3, buf4, 16, grid=grid(16), stream=stream0)
buf5 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [softmax], Original ATen: [aten._softmax]
triton_poi_fused__softmax_3.run(buf4, buf5, 16, grid=grid(16), stream=stream0)
buf6 = buf4; del buf4 # reuse
# Topologically Sorted Source Nodes: [mul, x_1], Original ATen: [aten.mul, aten.sum]
triton_poi_fused_mul_sum_4.run(primals_1, buf5, buf6, 16, grid=grid(16), stream=stream0)
del buf5
return (buf6, primals_1, reinterpret_tensor(buf0, (16, 4), (4, 1), 0), buf2, buf3, reinterpret_tensor(primals_4, (1, 4), (1, 1), 0), )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, 1), (1, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.nn.functional as F
class Self_Attentive_Pooling(nn.Module):
def __init__(self, dim):
"""SAP
Paper: Self-Attentive Speaker Embeddings for Text-Independent Speaker Verification
Link： https://danielpovey.com/files/2018_interspeech_xvector_attention.pdf
Args:
dim (pair): the size of attention weights
"""
super(Self_Attentive_Pooling, self).__init__()
self.sap_linear = nn.Linear(dim, dim)
self.attention = nn.Parameter(torch.FloatTensor(dim, 1))
def forward(self, x):
"""Computes Self-Attentive Pooling Module
Args:
x (torch.Tensor): Input tensor (#batch, dim, frames).
Returns:
torch.Tensor: Output tensor (#batch, dim)
"""
x = x.permute(0, 2, 1)
h = torch.tanh(self.sap_linear(x))
w = torch.matmul(h, self.attention).squeeze(dim=2)
w = F.softmax(w, dim=1).view(x.size(0), x.size(1), 1)
x = torch.sum(x * w, dim=1)
return x
def get_inputs():
return [torch.rand([4, 4, 4])]
def get_init_inputs():
return [[], {'dim': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_clone_0(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK: tl.
constexpr, XBLOCK: tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = yindex // 4
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + 4 * x2 + 16 * y1), xmask & ymask,
eviction_policy='evict_last')
tl.store(out_ptr0 + (x2 + 4 * y3), tmp0, xmask & ymask)
@triton.jit
def triton_poi_fused_add_tanh_1(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl.
constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = libdevice.tanh(tmp2)
tl.store(in_out_ptr0 + x2, tmp3, xmask)
@triton.jit
def triton_poi_fused__softmax_2(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp3 = triton_helpers.maximum(tmp1, tmp2)
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp7 = triton_helpers.maximum(tmp5, tmp6)
tmp8 = tmp0 - tmp7
tmp9 = tl_math.exp(tmp8)
tl.store(out_ptr0 + x2, tmp9, xmask)
@triton.jit
def triton_poi_fused__softmax_3(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + x2, tmp8, xmask)
@triton.jit
def triton_poi_fused_mul_sum_4(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK:
tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + 4 * x2, xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + 4 * x1, xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr0 + (1 + 4 * x2), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr1 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr0 + (2 + 4 * x2), xmask, eviction_policy='evict_last')
tmp8 = tl.load(in_ptr1 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp11 = tl.load(in_ptr0 + (3 + 4 * x2), xmask, eviction_policy='evict_last'
)
tmp12 = tl.load(in_ptr1 + (3 + 4 * x1), xmask, eviction_policy='evict_last'
)
tmp2 = tmp0 * tmp1
tmp5 = tmp3 * tmp4
tmp6 = tmp2 + tmp5
tmp9 = tmp7 * tmp8
tmp10 = tmp6 + tmp9
tmp13 = tmp11 * tmp12
tmp14 = tmp10 + tmp13
tl.store(out_ptr0 + x2, tmp14, xmask)
def call(args):
primals_1, primals_2, primals_3, primals_4 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_2, (4, 4), (4, 1))
assert_size_stride(primals_3, (4,), (1,))
assert_size_stride(primals_4, (4, 1), (1, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_clone_0[grid(16, 4)](primals_1, buf0, 16, 4,
XBLOCK=4, YBLOCK=16, num_warps=1, num_stages=1)
buf1 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(buf0, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_2, (4, 4), (1, 4), 0), out=buf1)
del primals_2
buf2 = reinterpret_tensor(buf1, (4, 4, 4), (16, 4, 1), 0)
del buf1
triton_poi_fused_add_tanh_1[grid(64)](buf2, primals_3, 64, XBLOCK=
64, num_warps=1, num_stages=1)
del primals_3
buf3 = empty_strided_cuda((16, 1), (1, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(buf2, (16, 4), (4, 1), 0),
primals_4, out=buf3)
buf4 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
triton_poi_fused__softmax_2[grid(16)](buf3, buf4, 16, XBLOCK=16,
num_warps=1, num_stages=1)
buf5 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
triton_poi_fused__softmax_3[grid(16)](buf4, buf5, 16, XBLOCK=16,
num_warps=1, num_stages=1)
buf6 = buf4
del buf4
triton_poi_fused_mul_sum_4[grid(16)](primals_1, buf5, buf6, 16,
XBLOCK=16, num_warps=1, num_stages=1)
del buf5
return buf6, primals_1, reinterpret_tensor(buf0, (16, 4), (4, 1), 0
), buf2, buf3, reinterpret_tensor(primals_4, (1, 4), (1, 1), 0)
class Self_Attentive_PoolingNew(nn.Module):
def __init__(self, dim):
"""SAP
Paper: Self-Attentive Speaker Embeddings for Text-Independent Speaker Verification
Link： https://danielpovey.com/files/2018_interspeech_xvector_attention.pdf
Args:
dim (pair): the size of attention weights
"""
super(Self_Attentive_PoolingNew, self).__init__()
self.sap_linear = nn.Linear(dim, dim)
self.attention = nn.Parameter(torch.FloatTensor(dim, 1))
def forward(self, input_0):
primals_4 = self.attention
primals_2 = self.sap_linear.weight
primals_3 = self.sap_linear.bias
primals_1 = input_0
output = call([primals_1, primals_2, primals_3, primals_4])
return output[0]
| ikunsaikou/lhy_ML2021Spring | Self_Attentive_Pooling | false | 6,870 | [
"WTFPL"
] | 1 | 80d8922077e2f5abba6a440c17654a143ebc8c9c | https://github.com/ikunsaikou/lhy_ML2021Spring/tree/80d8922077e2f5abba6a440c17654a143ebc8c9c | import torch
import torch.nn as nn
import torch.nn.functional as F
class Model(nn.Module):
def __init__(self, dim):
"""SAP
Paper: Self-Attentive Speaker Embeddings for Text-Independent Speaker Verification
Link： https://danielpovey.com/files/2018_interspeech_xvector_attention.pdf
Args:
dim (pair): the size of attention weights
"""
super().__init__()
self.sap_linear = nn.Linear(dim, dim)
self.attention = nn.Parameter(torch.FloatTensor(dim, 1))
def forward(self, x):
"""Computes Self-Attentive Pooling Module
Args:
x (torch.Tensor): Input tensor (#batch, dim, frames).
Returns:
torch.Tensor: Output tensor (#batch, dim)
"""
x = x.permute(0, 2, 1)
h = torch.tanh(self.sap_linear(x))
w = torch.matmul(h, self.attention).squeeze(dim=2)
w = F.softmax(w, dim=1).view(x.size(0), x.size(1), 1)
x = torch.sum(x * w, dim=1)
return x
def get_inputs():
return [torch.rand([4, 4, 4])]
def get_init_inputs():
return [4]
|
FCLateActionSAQFunction | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/lg/clgg7lidrsfgaud7ua7lulq4csdjjfhoa5iu6ldcldqvnqeoprch.py
# Topologically Sorted Source Nodes: [h_1], Original ATen: [aten.cat]
# Source node to ATen node mapping:
# h_1 => cat
# Graph fragment:
# %cat : [num_users=2] = call_function[target=torch.ops.aten.cat.default](args = ([%relu, %primals_4], 1), kwargs = {})
triton_poi_fused_cat_0 = async_compile.triton('triton_poi_fused_cat_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[32],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_cat_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 3, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_cat_0(in_ptr0, in_ptr1, in_ptr2, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 32
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 8
x1 = (xindex // 8)
x2 = xindex
tmp0 = x0
tmp1 = tl.full([1], 0, tl.int64)
tmp2 = tmp0 >= tmp1
tmp3 = tl.full([1], 4, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = tl.load(in_ptr0 + ((4*x1) + x0), tmp4 & xmask, eviction_policy='evict_last', other=0.0)
tmp6 = tl.load(in_ptr1 + (x0), tmp4 & xmask, eviction_policy='evict_last', other=0.0)
tmp7 = tmp5 + tmp6
tmp8 = tl.full([1], 0, tl.int32)
tmp9 = triton_helpers.maximum(tmp8, tmp7)
tmp10 = tl.full(tmp9.shape, 0.0, tmp9.dtype)
tmp11 = tl.where(tmp4, tmp9, tmp10)
tmp12 = tmp0 >= tmp3
tmp13 = tl.full([1], 8, tl.int64)
tmp14 = tmp0 < tmp13
tmp15 = tl.load(in_ptr2 + ((4*x1) + ((-4) + x0)), tmp12 & xmask, eviction_policy='evict_last', other=0.0)
tmp16 = tl.where(tmp4, tmp11, tmp15)
tl.store(out_ptr0 + (x2), tmp16, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/p5/cp5tfqfwrfwxlkkeq5u27dwcgidea534dshg3prbpaz2f7xapt3e.py
# Topologically Sorted Source Nodes: [h], Original ATen: [aten.relu, aten.threshold_backward]
# Source node to ATen node mapping:
# h => relu
# Graph fragment:
# %add_tensor : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mm_default, %primals_3), kwargs = {})
# %relu : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%add_tensor,), kwargs = {})
# %le : [num_users=1] = call_function[target=torch.ops.aten.le.Scalar](args = (%relu, 0), kwargs = {})
triton_poi_fused_relu_threshold_backward_1 = async_compile.triton('triton_poi_fused_relu_threshold_backward_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*i1', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_relu_threshold_backward_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_relu_threshold_backward_1(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr1 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(out_ptr0 + (x2), tmp6, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6 = args
args.clear()
assert_size_stride(primals_1, (4, 4), (4, 1))
assert_size_stride(primals_2, (4, 4), (4, 1))
assert_size_stride(primals_3, (4, ), (1, ))
assert_size_stride(primals_4, (4, 4), (4, 1))
assert_size_stride(primals_5, (1, 8), (8, 1))
assert_size_stride(primals_6, (1, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(primals_1, reinterpret_tensor(primals_2, (4, 4), (1, 4), 0), out=buf0)
del primals_2
buf1 = empty_strided_cuda((4, 8), (8, 1), torch.float32)
# Topologically Sorted Source Nodes: [h_1], Original ATen: [aten.cat]
stream0 = get_raw_stream(0)
triton_poi_fused_cat_0.run(buf0, primals_3, primals_4, buf1, 32, grid=grid(32), stream=stream0)
del primals_4
buf3 = empty_strided_cuda((4, 1), (1, 1), torch.float32)
# Topologically Sorted Source Nodes: [linear_1], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_6, buf1, reinterpret_tensor(primals_5, (8, 1), (1, 8), 0), alpha=1, beta=1, out=buf3)
del primals_6
buf4 = empty_strided_cuda((4, 4), (4, 1), torch.bool)
# Topologically Sorted Source Nodes: [h], Original ATen: [aten.relu, aten.threshold_backward]
triton_poi_fused_relu_threshold_backward_1.run(buf0, primals_3, buf4, 16, grid=grid(16), stream=stream0)
del buf0
del primals_3
return (buf3, primals_1, buf1, primals_5, buf4, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((1, 8), (8, 1), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((1, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import numpy as np
import torch.nn as nn
import torch.nn.functional as F
from abc import ABCMeta
from abc import abstractmethod
def init_lecun_normal(tensor, scale=1.0):
"""Initializes the tensor with LeCunNormal."""
fan_in = torch.nn.init._calculate_correct_fan(tensor, 'fan_in')
std = scale * np.sqrt(1.0 / fan_in)
with torch.no_grad():
return tensor.normal_(0, std)
@torch.no_grad()
def init_chainer_default(layer):
"""Initializes the layer with the chainer default.
weights with LeCunNormal(scale=1.0) and zeros as biases
"""
assert isinstance(layer, nn.Module)
if isinstance(layer, (nn.Linear, nn.Conv2d)):
init_lecun_normal(layer.weight)
if layer.bias is not None:
nn.init.zeros_(layer.bias)
return layer
class MLP(nn.Module):
"""Multi-Layer Perceptron"""
def __init__(self, in_size, out_size, hidden_sizes, nonlinearity=F.relu,
last_wscale=1):
self.in_size = in_size
self.out_size = out_size
self.hidden_sizes = hidden_sizes
self.nonlinearity = nonlinearity
super().__init__()
if hidden_sizes:
self.hidden_layers = nn.ModuleList()
self.hidden_layers.append(nn.Linear(in_size, hidden_sizes[0]))
for hin, hout in zip(hidden_sizes, hidden_sizes[1:]):
self.hidden_layers.append(nn.Linear(hin, hout))
self.hidden_layers.apply(init_chainer_default)
self.output = nn.Linear(hidden_sizes[-1], out_size)
else:
self.output = nn.Linear(in_size, out_size)
init_lecun_normal(self.output.weight, scale=last_wscale)
nn.init.zeros_(self.output.bias)
def forward(self, x):
h = x
if self.hidden_sizes:
for l in self.hidden_layers:
h = self.nonlinearity(l(h))
return self.output(h)
class StateActionQFunction(object, metaclass=ABCMeta):
"""Abstract Q-function with state and action input."""
@abstractmethod
def __call__(self, x, a):
"""Evaluates Q-function
Args:
x (ndarray): state input
a (ndarray): action input
Returns:
Q-value for state x and action a
"""
raise NotImplementedError()
class FCLateActionSAQFunction(nn.Module, StateActionQFunction):
"""Fully-connected (s,a)-input Q-function with late action input.
Actions are not included until the second hidden layer and not normalized.
This architecture is used in the DDPG paper:
http://arxiv.org/abs/1509.02971
Args:
n_dim_obs (int): Number of dimensions of observation space.
n_dim_action (int): Number of dimensions of action space.
n_hidden_channels (int): Number of hidden channels.
n_hidden_layers (int): Number of hidden layers. It must be greater than
or equal to 1.
nonlinearity (callable): Nonlinearity between layers. It must accept a
Variable as an argument and return a Variable with the same shape.
Nonlinearities with learnable parameters such as PReLU are not
supported.
last_wscale (float): Scale of weight initialization of the last layer.
"""
def __init__(self, n_dim_obs, n_dim_action, n_hidden_channels,
n_hidden_layers, nonlinearity=F.relu, last_wscale=1.0):
assert n_hidden_layers >= 1
self.n_input_channels = n_dim_obs + n_dim_action
self.n_hidden_layers = n_hidden_layers
self.n_hidden_channels = n_hidden_channels
self.nonlinearity = nonlinearity
super().__init__()
self.obs_mlp = MLP(in_size=n_dim_obs, out_size=n_hidden_channels,
hidden_sizes=[])
self.mlp = MLP(in_size=n_hidden_channels + n_dim_action, out_size=1,
hidden_sizes=[self.n_hidden_channels] * (self.n_hidden_layers -
1), nonlinearity=nonlinearity, last_wscale=last_wscale)
self.output = self.mlp.output
def forward(self, state, action):
h = self.nonlinearity(self.obs_mlp(state))
h = torch.cat((h, action), dim=1)
return self.mlp(h)
def get_inputs():
return [torch.rand([4, 4]), torch.rand([4, 4])]
def get_init_inputs():
return [[], {'n_dim_obs': 4, 'n_dim_action': 4, 'n_hidden_channels': 4,
'n_hidden_layers': 1}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
import numpy as np
import torch.nn as nn
import torch.nn.functional as F
from abc import ABCMeta
from abc import abstractmethod
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_cat_0(in_ptr0, in_ptr1, in_ptr2, out_ptr0, xnumel,
XBLOCK: tl.constexpr):
xnumel = 32
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 8
x1 = xindex // 8
x2 = xindex
tmp0 = x0
tl.full([1], 0, tl.int64)
tmp3 = tl.full([1], 4, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = tl.load(in_ptr0 + (4 * x1 + x0), tmp4 & xmask, eviction_policy=
'evict_last', other=0.0)
tmp6 = tl.load(in_ptr1 + x0, tmp4 & xmask, eviction_policy='evict_last',
other=0.0)
tmp7 = tmp5 + tmp6
tmp8 = tl.full([1], 0, tl.int32)
tmp9 = triton_helpers.maximum(tmp8, tmp7)
tmp10 = tl.full(tmp9.shape, 0.0, tmp9.dtype)
tmp11 = tl.where(tmp4, tmp9, tmp10)
tmp12 = tmp0 >= tmp3
tl.full([1], 8, tl.int64)
tmp15 = tl.load(in_ptr2 + (4 * x1 + (-4 + x0)), tmp12 & xmask,
eviction_policy='evict_last', other=0.0)
tmp16 = tl.where(tmp4, tmp11, tmp15)
tl.store(out_ptr0 + x2, tmp16, xmask)
@triton.jit
def triton_poi_fused_relu_threshold_backward_1(in_ptr0, in_ptr1, out_ptr0,
xnumel, XBLOCK: tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr1 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(out_ptr0 + x2, tmp6, xmask)
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6 = args
args.clear()
assert_size_stride(primals_1, (4, 4), (4, 1))
assert_size_stride(primals_2, (4, 4), (4, 1))
assert_size_stride(primals_3, (4,), (1,))
assert_size_stride(primals_4, (4, 4), (4, 1))
assert_size_stride(primals_5, (1, 8), (8, 1))
assert_size_stride(primals_6, (1,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.mm(primals_1, reinterpret_tensor(primals_2, (4, 4),
(1, 4), 0), out=buf0)
del primals_2
buf1 = empty_strided_cuda((4, 8), (8, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_cat_0[grid(32)](buf0, primals_3, primals_4, buf1,
32, XBLOCK=32, num_warps=1, num_stages=1)
del primals_4
buf3 = empty_strided_cuda((4, 1), (1, 1), torch.float32)
extern_kernels.addmm(primals_6, buf1, reinterpret_tensor(primals_5,
(8, 1), (1, 8), 0), alpha=1, beta=1, out=buf3)
del primals_6
buf4 = empty_strided_cuda((4, 4), (4, 1), torch.bool)
triton_poi_fused_relu_threshold_backward_1[grid(16)](buf0,
primals_3, buf4, 16, XBLOCK=16, num_warps=1, num_stages=1)
del buf0
del primals_3
return buf3, primals_1, buf1, primals_5, buf4
def init_lecun_normal(tensor, scale=1.0):
"""Initializes the tensor with LeCunNormal."""
fan_in = torch.nn.init._calculate_correct_fan(tensor, 'fan_in')
std = scale * np.sqrt(1.0 / fan_in)
with torch.no_grad():
return tensor.normal_(0, std)
@torch.no_grad()
def init_chainer_default(layer):
"""Initializes the layer with the chainer default.
weights with LeCunNormal(scale=1.0) and zeros as biases
"""
assert isinstance(layer, nn.Module)
if isinstance(layer, (nn.Linear, nn.Conv2d)):
init_lecun_normal(layer.weight)
if layer.bias is not None:
nn.init.zeros_(layer.bias)
return layer
class MLP(nn.Module):
"""Multi-Layer Perceptron"""
def __init__(self, in_size, out_size, hidden_sizes, nonlinearity=F.relu,
last_wscale=1):
self.in_size = in_size
self.out_size = out_size
self.hidden_sizes = hidden_sizes
self.nonlinearity = nonlinearity
super().__init__()
if hidden_sizes:
self.hidden_layers = nn.ModuleList()
self.hidden_layers.append(nn.Linear(in_size, hidden_sizes[0]))
for hin, hout in zip(hidden_sizes, hidden_sizes[1:]):
self.hidden_layers.append(nn.Linear(hin, hout))
self.hidden_layers.apply(init_chainer_default)
self.output = nn.Linear(hidden_sizes[-1], out_size)
else:
self.output = nn.Linear(in_size, out_size)
init_lecun_normal(self.output.weight, scale=last_wscale)
nn.init.zeros_(self.output.bias)
def forward(self, x):
h = x
if self.hidden_sizes:
for l in self.hidden_layers:
h = self.nonlinearity(l(h))
return self.output(h)
class StateActionQFunction(object, metaclass=ABCMeta):
"""Abstract Q-function with state and action input."""
@abstractmethod
def __call__(self, x, a):
"""Evaluates Q-function
Args:
x (ndarray): state input
a (ndarray): action input
Returns:
Q-value for state x and action a
"""
raise NotImplementedError()
class FCLateActionSAQFunctionNew(nn.Module, StateActionQFunction):
"""Fully-connected (s,a)-input Q-function with late action input.
Actions are not included until the second hidden layer and not normalized.
This architecture is used in the DDPG paper:
http://arxiv.org/abs/1509.02971
Args:
n_dim_obs (int): Number of dimensions of observation space.
n_dim_action (int): Number of dimensions of action space.
n_hidden_channels (int): Number of hidden channels.
n_hidden_layers (int): Number of hidden layers. It must be greater than
or equal to 1.
nonlinearity (callable): Nonlinearity between layers. It must accept a
Variable as an argument and return a Variable with the same shape.
Nonlinearities with learnable parameters such as PReLU are not
supported.
last_wscale (float): Scale of weight initialization of the last layer.
"""
def __init__(self, n_dim_obs, n_dim_action, n_hidden_channels,
n_hidden_layers, nonlinearity=F.relu, last_wscale=1.0):
assert n_hidden_layers >= 1
self.n_input_channels = n_dim_obs + n_dim_action
self.n_hidden_layers = n_hidden_layers
self.n_hidden_channels = n_hidden_channels
self.nonlinearity = nonlinearity
super().__init__()
self.obs_mlp = MLP(in_size=n_dim_obs, out_size=n_hidden_channels,
hidden_sizes=[])
self.mlp = MLP(in_size=n_hidden_channels + n_dim_action, out_size=1,
hidden_sizes=[self.n_hidden_channels] * (self.n_hidden_layers -
1), nonlinearity=nonlinearity, last_wscale=last_wscale)
self.output = self.mlp.output
def forward(self, input_0, input_1):
primals_1 = self.obs_mlp.output.weight
primals_3 = self.obs_mlp.output.bias
primals_5 = self.mlp.output.weight
primals_6 = self.mlp.output.bias
primals_2 = input_0
primals_4 = input_1
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6])
return output[0]
| imatge-upc/pixelcoordEDL | FCLateActionSAQFunction | false | 6,871 | [
"MIT"
] | 1 | 353632feed6ac8c93758c1a2a1b7a477e7ff053c | https://github.com/imatge-upc/pixelcoordEDL/tree/353632feed6ac8c93758c1a2a1b7a477e7ff053c | import torch
import numpy as np
import torch.nn as nn
import torch.nn.functional as F
from abc import ABCMeta
from abc import abstractmethod
def init_lecun_normal(tensor, scale=1.0):
"""Initializes the tensor with LeCunNormal."""
fan_in = torch.nn.init._calculate_correct_fan(tensor, 'fan_in')
std = scale * np.sqrt(1.0 / fan_in)
with torch.no_grad():
return tensor.normal_(0, std)
@torch.no_grad()
def init_chainer_default(layer):
"""Initializes the layer with the chainer default.
weights with LeCunNormal(scale=1.0) and zeros as biases
"""
assert isinstance(layer, nn.Module)
if isinstance(layer, (nn.Linear, nn.Conv2d)):
init_lecun_normal(layer.weight)
if layer.bias is not None:
nn.init.zeros_(layer.bias)
return layer
class MLP(nn.Module):
"""Multi-Layer Perceptron"""
def __init__(self, in_size, out_size, hidden_sizes, nonlinearity=F.relu,
last_wscale=1):
self.in_size = in_size
self.out_size = out_size
self.hidden_sizes = hidden_sizes
self.nonlinearity = nonlinearity
super().__init__()
if hidden_sizes:
self.hidden_layers = nn.ModuleList()
self.hidden_layers.append(nn.Linear(in_size, hidden_sizes[0]))
for hin, hout in zip(hidden_sizes, hidden_sizes[1:]):
self.hidden_layers.append(nn.Linear(hin, hout))
self.hidden_layers.apply(init_chainer_default)
self.output = nn.Linear(hidden_sizes[-1], out_size)
else:
self.output = nn.Linear(in_size, out_size)
init_lecun_normal(self.output.weight, scale=last_wscale)
nn.init.zeros_(self.output.bias)
def forward(self, x):
h = x
if self.hidden_sizes:
for l in self.hidden_layers:
h = self.nonlinearity(l(h))
return self.output(h)
class StateActionQFunction(object, metaclass=ABCMeta):
"""Abstract Q-function with state and action input."""
@abstractmethod
def __call__(self, x, a):
"""Evaluates Q-function
Args:
x (ndarray): state input
a (ndarray): action input
Returns:
Q-value for state x and action a
"""
raise NotImplementedError()
class Model(nn.Module, StateActionQFunction):
"""Fully-connected (s,a)-input Q-function with late action input.
Actions are not included until the second hidden layer and not normalized.
This architecture is used in the DDPG paper:
http://arxiv.org/abs/1509.02971
Args:
n_dim_obs (int): Number of dimensions of observation space.
n_dim_action (int): Number of dimensions of action space.
n_hidden_channels (int): Number of hidden channels.
n_hidden_layers (int): Number of hidden layers. It must be greater than
or equal to 1.
nonlinearity (callable): Nonlinearity between layers. It must accept a
Variable as an argument and return a Variable with the same shape.
Nonlinearities with learnable parameters such as PReLU are not
supported.
last_wscale (float): Scale of weight initialization of the last layer.
"""
def __init__(self, n_dim_obs, n_dim_action, n_hidden_channels,
n_hidden_layers, nonlinearity=F.relu, last_wscale=1.0):
assert n_hidden_layers >= 1
self.n_input_channels = n_dim_obs + n_dim_action
self.n_hidden_layers = n_hidden_layers
self.n_hidden_channels = n_hidden_channels
self.nonlinearity = nonlinearity
super().__init__()
self.obs_mlp = MLP(in_size=n_dim_obs, out_size=n_hidden_channels,
hidden_sizes=[])
self.mlp = MLP(in_size=n_hidden_channels + n_dim_action, out_size=1,
hidden_sizes=[self.n_hidden_channels] * (self.n_hidden_layers -
1), nonlinearity=nonlinearity, last_wscale=last_wscale)
self.output = self.mlp.output
def f
# ... truncated (>4000 chars) for memory efficiency |
TransformerEncoderLayer | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/qw/cqw7yoyglmtjad3kirznl5odetqfs3k6pjtnfdbzklyhsdvuvgft.py
# Topologically Sorted Source Nodes: [multi_head_attention_forward], Original ATen: [aten.mul]
# Source node to ATen node mapping:
# multi_head_attention_forward => mul
# Graph fragment:
# %mul : [num_users=2] = call_function[target=torch.ops.aten.mul.Tensor](args = (%permute_3, 1.0), kwargs = {})
triton_poi_fused_mul_0 = async_compile.triton('triton_poi_fused_mul_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_mul_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_mul_0(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = 1.0
tmp4 = tmp2 * tmp3
tl.store(in_out_ptr0 + (x2), tmp4, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/hz/chzi3aam26mikdhljz5x7jlqazm7kpktzeptsf36thgfhsg7ub6a.py
# Topologically Sorted Source Nodes: [multi_head_attention_forward], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# multi_head_attention_forward => amax, exp, sub
# Graph fragment:
# %amax : [num_users=1] = call_function[target=torch.ops.aten.amax.default](args = (%bmm, [-1], True), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%bmm, %amax), kwargs = {})
# %exp : [num_users=2] = call_function[target=torch.ops.aten.exp.default](args = (%sub,), kwargs = {})
triton_poi_fused__softmax_1 = async_compile.triton('triton_poi_fused__softmax_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_1(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp3 = triton_helpers.maximum(tmp1, tmp2)
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp7 = triton_helpers.maximum(tmp5, tmp6)
tmp8 = tmp0 - tmp7
tmp9 = tl_math.exp(tmp8)
tl.store(out_ptr0 + (x2), tmp9, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/em/cem6qbxwbiqnjqybzk5arf2obt5uggy4qs7otwwpovvnrhvdc6h4.py
# Topologically Sorted Source Nodes: [multi_head_attention_forward], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# multi_head_attention_forward => div, sum_1
# Graph fragment:
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%exp, [-1], True), kwargs = {})
# %div : [num_users=2] = call_function[target=torch.ops.aten.div.Tensor](args = (%exp, %sum_1), kwargs = {})
triton_poi_fused__softmax_2 = async_compile.triton('triton_poi_fused__softmax_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_2(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + (x2), tmp8, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/rh/crhjfwyl6xoj5ylcsbbh6lp2vlegits2zkdej3b3wb2q4ddfnejv.py
# Topologically Sorted Source Nodes: [multi_head_attention_forward], Original ATen: [aten.clone]
# Source node to ATen node mapping:
# multi_head_attention_forward => clone
# Graph fragment:
# %clone : [num_users=1] = call_function[target=torch.ops.aten.clone.default](args = (%permute_7,), kwargs = {memory_format: torch.contiguous_format})
triton_poi_fused_clone_3 = async_compile.triton('triton_poi_fused_clone_3', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[4, 4], tile_hint=TileHint.SQUARE,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clone_3', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clone_3(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 4
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x1 = xindex
y0 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + (4*x1)), xmask & ymask)
tl.store(out_ptr0 + (x1 + (4*y0)), tmp0, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/5z/c5zy7julai2lhuinuwjgyl62nx7cyws6ni5poe5jzp7qn532rcgh.py
# Topologically Sorted Source Nodes: [src], Original ATen: [aten.add]
# Source node to ATen node mapping:
# src => add
# Graph fragment:
# %add : [num_users=3] = call_function[target=torch.ops.aten.add.Tensor](args = (%primals_1, %squeeze), kwargs = {})
triton_poi_fused_add_4 = async_compile.triton('triton_poi_fused_add_4', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_4', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 3, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_4(in_out_ptr0, in_ptr0, in_ptr1, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_out_ptr0 + (x2), xmask)
tmp2 = tl.load(in_ptr1 + (x0), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp4 = tmp0 + tmp3
tl.store(in_out_ptr0 + (x2), tmp4, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/p5/cp5zfsohdolod3ikxheo5zxosgk7sw3fpun4wu4lt34g7uwo2pdt.py
# Topologically Sorted Source Nodes: [leaky_relu], Original ATen: [aten.leaky_relu]
# Source node to ATen node mapping:
# leaky_relu => gt, mul_1, where
# Graph fragment:
# %add_tensor_1 : [num_users=3] = call_function[target=torch.ops.aten.add.Tensor](args = (%mm_default_1, %primals_7), kwargs = {})
# %gt : [num_users=2] = call_function[target=torch.ops.aten.gt.Scalar](args = (%add_tensor_1, 0), kwargs = {})
# %mul_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%add_tensor_1, True), kwargs = {})
# %where : [num_users=2] = call_function[target=torch.ops.aten.where.self](args = (%gt, %add_tensor_1, %mul_1), kwargs = {})
triton_poi_fused_leaky_relu_5 = async_compile.triton('triton_poi_fused_leaky_relu_5', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*i1', 3: '*fp32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_leaky_relu_5', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_leaky_relu_5(in_ptr0, in_ptr1, out_ptr0, out_ptr1, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 16
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr1 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = 0.0
tmp4 = tmp2 > tmp3
tmp5 = 1.0
tmp6 = tmp2 * tmp5
tmp7 = tl.where(tmp4, tmp2, tmp6)
tl.store(out_ptr0 + (x2), tmp4, xmask)
tl.store(out_ptr1 + (x2), tmp7, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9 = args
args.clear()
assert_size_stride(primals_1, (4, 4), (4, 1))
assert_size_stride(primals_2, (12, 4), (4, 1))
assert_size_stride(primals_3, (12, ), (1, ))
assert_size_stride(primals_4, (4, 4), (4, 1))
assert_size_stride(primals_5, (4, ), (1, ))
assert_size_stride(primals_6, (16, 4), (4, 1))
assert_size_stride(primals_7, (16, ), (1, ))
assert_size_stride(primals_8, (4, 16), (16, 1))
assert_size_stride(primals_9, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(primals_1, reinterpret_tensor(primals_2, (4, 4), (1, 4), 0), out=buf0)
buf1 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [multi_head_attention_forward], Original ATen: [aten.addmm]
extern_kernels.addmm(reinterpret_tensor(primals_3, (4, ), (1, ), 4), primals_1, reinterpret_tensor(primals_2, (4, 4), (1, 4), 16), alpha=1, beta=1, out=buf1)
buf2 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [multi_head_attention_forward], Original ATen: [aten.addmm]
extern_kernels.addmm(reinterpret_tensor(primals_3, (4, ), (1, ), 8), primals_1, reinterpret_tensor(primals_2, (4, 4), (1, 4), 32), alpha=1, beta=1, out=buf2)
del primals_2
buf3 = reinterpret_tensor(buf0, (4, 4, 1), (1, 4, 16), 0); del buf0 # reuse
# Topologically Sorted Source Nodes: [multi_head_attention_forward], Original ATen: [aten.mul]
stream0 = get_raw_stream(0)
triton_poi_fused_mul_0.run(buf3, primals_3, 16, grid=grid(16), stream=stream0)
del primals_3
buf4 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [multi_head_attention_forward], Original ATen: [aten.bmm]
extern_kernels.bmm(buf3, reinterpret_tensor(buf1, (4, 1, 4), (1, 1, 4), 0), out=buf4)
buf5 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [multi_head_attention_forward], Original ATen: [aten._softmax]
triton_poi_fused__softmax_1.run(buf4, buf5, 64, grid=grid(64), stream=stream0)
buf6 = buf4; del buf4 # reuse
# Topologically Sorted Source Nodes: [multi_head_attention_forward], Original ATen: [aten._softmax]
triton_poi_fused__softmax_2.run(buf5, buf6, 64, grid=grid(64), stream=stream0)
buf7 = empty_strided_cuda((4, 4, 1), (4, 1, 1), torch.float32)
# Topologically Sorted Source Nodes: [multi_head_attention_forward], Original ATen: [aten.bmm]
extern_kernels.bmm(buf6, reinterpret_tensor(buf2, (4, 4, 1), (1, 4, 1), 0), out=buf7)
buf8 = empty_strided_cuda((4, 4, 1), (4, 1, 1), torch.float32)
# Topologically Sorted Source Nodes: [multi_head_attention_forward], Original ATen: [aten.clone]
triton_poi_fused_clone_3.run(buf7, buf8, 4, 4, grid=grid(4, 4), stream=stream0)
buf9 = reinterpret_tensor(buf7, (4, 4), (4, 1), 0); del buf7 # reuse
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(buf8, (4, 4), (4, 1), 0), reinterpret_tensor(primals_4, (4, 4), (1, 4), 0), out=buf9)
buf10 = buf9; del buf9 # reuse
# Topologically Sorted Source Nodes: [src], Original ATen: [aten.add]
triton_poi_fused_add_4.run(buf10, primals_1, primals_5, 16, grid=grid(16), stream=stream0)
del primals_5
buf11 = reinterpret_tensor(buf5, (4, 16), (16, 1), 0); del buf5 # reuse
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(buf10, reinterpret_tensor(primals_6, (4, 16), (1, 4), 0), out=buf11)
buf12 = empty_strided_cuda((4, 16), (16, 1), torch.bool)
buf13 = empty_strided_cuda((4, 16), (16, 1), torch.float32)
# Topologically Sorted Source Nodes: [leaky_relu], Original ATen: [aten.leaky_relu]
triton_poi_fused_leaky_relu_5.run(buf11, primals_7, buf12, buf13, 64, grid=grid(64), stream=stream0)
del buf11
del primals_7
buf14 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(buf13, reinterpret_tensor(primals_8, (16, 4), (1, 16), 0), out=buf14)
buf15 = buf14; del buf14 # reuse
# Topologically Sorted Source Nodes: [src_1], Original ATen: [aten.add]
triton_poi_fused_add_4.run(buf15, buf10, primals_9, 16, grid=grid(16), stream=stream0)
del primals_9
return (buf15, primals_1, buf6, reinterpret_tensor(buf8, (4, 4), (4, 1), 0), buf10, buf12, buf13, primals_8, primals_6, primals_4, reinterpret_tensor(buf2, (4, 1, 4), (1, 1, 4), 0), reinterpret_tensor(buf3, (4, 1, 4), (1, 1, 4), 0), reinterpret_tensor(buf1, (4, 4, 1), (1, 4, 1), 0), )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((12, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((12, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((16, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((16, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_8 = rand_strided((4, 16), (16, 1), device='cuda:0', dtype=torch.float32)
primals_9 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
class TransformerEncoderLayer(nn.Module):
def __init__(self, d_model, nhead, dim_feedforward=16, dropout=0):
super(TransformerEncoderLayer, self).__init__()
self.self_attn = nn.MultiheadAttention(d_model, nhead, dropout=dropout)
self.linear1 = nn.Linear(d_model, dim_feedforward)
self.dropout = nn.Dropout(dropout)
self.linear2 = nn.Linear(dim_feedforward, d_model)
self.dropout1 = nn.Dropout(dropout)
self.dropout2 = nn.Dropout(dropout)
self.activation = nn.LeakyReLU(True)
def forward(self, src, src_mask=None, src_key_padding_mask=None):
src2 = self.self_attn(src, src, src)[0]
src = src + self.dropout1(src2)
src2 = self.linear2(self.dropout(self.activation(self.linear1(src))))
src = src + self.dropout2(src2)
return src
def get_inputs():
return [torch.rand([4, 4])]
def get_init_inputs():
return [[], {'d_model': 4, 'nhead': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import math as tl_math
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_mul_0(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = 1.0
tmp4 = tmp2 * tmp3
tl.store(in_out_ptr0 + x2, tmp4, xmask)
@triton.jit
def triton_poi_fused__softmax_1(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp3 = triton_helpers.maximum(tmp1, tmp2)
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp7 = triton_helpers.maximum(tmp5, tmp6)
tmp8 = tmp0 - tmp7
tmp9 = tl_math.exp(tmp8)
tl.store(out_ptr0 + x2, tmp9, xmask)
@triton.jit
def triton_poi_fused__softmax_2(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + x2, tmp8, xmask)
@triton.jit
def triton_poi_fused_clone_3(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK: tl.
constexpr, XBLOCK: tl.constexpr):
ynumel = 4
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x1 = xindex
y0 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + 4 * x1), xmask & ymask)
tl.store(out_ptr0 + (x1 + 4 * y0), tmp0, xmask & ymask)
@triton.jit
def triton_poi_fused_add_4(in_out_ptr0, in_ptr0, in_ptr1, xnumel, XBLOCK:
tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_out_ptr0 + x2, xmask)
tmp2 = tl.load(in_ptr1 + x0, xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp4 = tmp0 + tmp3
tl.store(in_out_ptr0 + x2, tmp4, xmask)
@triton.jit
def triton_poi_fused_leaky_relu_5(in_ptr0, in_ptr1, out_ptr0, out_ptr1,
xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 16
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr1 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = 0.0
tmp4 = tmp2 > tmp3
tmp5 = 1.0
tmp6 = tmp2 * tmp5
tmp7 = tl.where(tmp4, tmp2, tmp6)
tl.store(out_ptr0 + x2, tmp4, xmask)
tl.store(out_ptr1 + x2, tmp7, xmask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7, primals_8, primals_9) = args
args.clear()
assert_size_stride(primals_1, (4, 4), (4, 1))
assert_size_stride(primals_2, (12, 4), (4, 1))
assert_size_stride(primals_3, (12,), (1,))
assert_size_stride(primals_4, (4, 4), (4, 1))
assert_size_stride(primals_5, (4,), (1,))
assert_size_stride(primals_6, (16, 4), (4, 1))
assert_size_stride(primals_7, (16,), (1,))
assert_size_stride(primals_8, (4, 16), (16, 1))
assert_size_stride(primals_9, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.mm(primals_1, reinterpret_tensor(primals_2, (4, 4),
(1, 4), 0), out=buf0)
buf1 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.addmm(reinterpret_tensor(primals_3, (4,), (1,), 4),
primals_1, reinterpret_tensor(primals_2, (4, 4), (1, 4), 16),
alpha=1, beta=1, out=buf1)
buf2 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.addmm(reinterpret_tensor(primals_3, (4,), (1,), 8),
primals_1, reinterpret_tensor(primals_2, (4, 4), (1, 4), 32),
alpha=1, beta=1, out=buf2)
del primals_2
buf3 = reinterpret_tensor(buf0, (4, 4, 1), (1, 4, 16), 0)
del buf0
get_raw_stream(0)
triton_poi_fused_mul_0[grid(16)](buf3, primals_3, 16, XBLOCK=16,
num_warps=1, num_stages=1)
del primals_3
buf4 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
extern_kernels.bmm(buf3, reinterpret_tensor(buf1, (4, 1, 4), (1, 1,
4), 0), out=buf4)
buf5 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
triton_poi_fused__softmax_1[grid(64)](buf4, buf5, 64, XBLOCK=64,
num_warps=1, num_stages=1)
buf6 = buf4
del buf4
triton_poi_fused__softmax_2[grid(64)](buf5, buf6, 64, XBLOCK=64,
num_warps=1, num_stages=1)
buf7 = empty_strided_cuda((4, 4, 1), (4, 1, 1), torch.float32)
extern_kernels.bmm(buf6, reinterpret_tensor(buf2, (4, 4, 1), (1, 4,
1), 0), out=buf7)
buf8 = empty_strided_cuda((4, 4, 1), (4, 1, 1), torch.float32)
triton_poi_fused_clone_3[grid(4, 4)](buf7, buf8, 4, 4, XBLOCK=4,
YBLOCK=4, num_warps=1, num_stages=1)
buf9 = reinterpret_tensor(buf7, (4, 4), (4, 1), 0)
del buf7
extern_kernels.mm(reinterpret_tensor(buf8, (4, 4), (4, 1), 0),
reinterpret_tensor(primals_4, (4, 4), (1, 4), 0), out=buf9)
buf10 = buf9
del buf9
triton_poi_fused_add_4[grid(16)](buf10, primals_1, primals_5, 16,
XBLOCK=16, num_warps=1, num_stages=1)
del primals_5
buf11 = reinterpret_tensor(buf5, (4, 16), (16, 1), 0)
del buf5
extern_kernels.mm(buf10, reinterpret_tensor(primals_6, (4, 16), (1,
4), 0), out=buf11)
buf12 = empty_strided_cuda((4, 16), (16, 1), torch.bool)
buf13 = empty_strided_cuda((4, 16), (16, 1), torch.float32)
triton_poi_fused_leaky_relu_5[grid(64)](buf11, primals_7, buf12,
buf13, 64, XBLOCK=64, num_warps=1, num_stages=1)
del buf11
del primals_7
buf14 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.mm(buf13, reinterpret_tensor(primals_8, (16, 4), (1,
16), 0), out=buf14)
buf15 = buf14
del buf14
triton_poi_fused_add_4[grid(16)](buf15, buf10, primals_9, 16,
XBLOCK=16, num_warps=1, num_stages=1)
del primals_9
return (buf15, primals_1, buf6, reinterpret_tensor(buf8, (4, 4), (4, 1),
0), buf10, buf12, buf13, primals_8, primals_6, primals_4,
reinterpret_tensor(buf2, (4, 1, 4), (1, 1, 4), 0),
reinterpret_tensor(buf3, (4, 1, 4), (1, 1, 4), 0),
reinterpret_tensor(buf1, (4, 4, 1), (1, 4, 1), 0))
class TransformerEncoderLayerNew(nn.Module):
def __init__(self, d_model, nhead, dim_feedforward=16, dropout=0):
super(TransformerEncoderLayerNew, self).__init__()
self.self_attn = nn.MultiheadAttention(d_model, nhead, dropout=dropout)
self.linear1 = nn.Linear(d_model, dim_feedforward)
self.dropout = nn.Dropout(dropout)
self.linear2 = nn.Linear(dim_feedforward, d_model)
self.dropout1 = nn.Dropout(dropout)
self.dropout2 = nn.Dropout(dropout)
self.activation = nn.LeakyReLU(True)
def forward(self, input_0):
primals_2 = self.self_attn.in_proj_weight
primals_3 = self.self_attn.in_proj_bias
primals_1 = self.self_attn.out_proj.weight
primals_5 = self.self_attn.out_proj.bias
primals_6 = self.linear1.weight
primals_7 = self.linear1.bias
primals_8 = self.linear2.weight
primals_9 = self.linear2.bias
primals_4 = input_0
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7, primals_8, primals_9])
return output[0]
| imperial-qore/CAROL | TransformerEncoderLayer | false | 6,872 | [
"BSD-3-Clause"
] | 1 | 57dc42c4ddeb9e75eed43a91ceb336a1ecc9c8b9 | https://github.com/imperial-qore/CAROL/tree/57dc42c4ddeb9e75eed43a91ceb336a1ecc9c8b9 | import torch
import torch.nn as nn
class Model(nn.Module):
def __init__(self, d_model, nhead, dim_feedforward=16, dropout=0):
super().__init__()
self.self_attn = nn.MultiheadAttention(d_model, nhead, dropout=dropout)
self.linear1 = nn.Linear(d_model, dim_feedforward)
self.dropout = nn.Dropout(dropout)
self.linear2 = nn.Linear(dim_feedforward, d_model)
self.dropout1 = nn.Dropout(dropout)
self.dropout2 = nn.Dropout(dropout)
self.activation = nn.LeakyReLU(True)
def forward(self, src, src_mask=None, src_key_padding_mask=None):
src2 = self.self_attn(src, src, src)[0]
src = src + self.dropout1(src2)
src2 = self.linear2(self.dropout(self.activation(self.linear1(src))))
src = src + self.dropout2(src2)
return src
def get_inputs():
return [torch.rand([4, 4])]
def get_init_inputs():
return [4, 4]
|
WNConv2d | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/4m/c4munax4bhqei64mhriszwqd42q6bjyh4nv3jazxhymu3f7wtucw.py
# Topologically Sorted Source Nodes: [_weight_norm], Original ATen: [aten._weight_norm_interface]
# Source node to ATen node mapping:
# _weight_norm => div, mul, pow_1, pow_2, sum_1
# Graph fragment:
# %pow_1 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%primals_2, 2), kwargs = {})
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%pow_1, [1, 2, 3], True), kwargs = {})
# %pow_2 : [num_users=2] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%sum_1, 0.5), kwargs = {})
# %div : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%primals_1, %pow_2), kwargs = {})
# %mul : [num_users=2] = call_function[target=torch.ops.aten.mul.Tensor](args = (%primals_2, %div), kwargs = {})
triton_per_fused__weight_norm_interface_0 = async_compile.triton('triton_per_fused__weight_norm_interface_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[4, 64],
reduction_hint=ReductionHint.INNER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32', 5: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 5), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused__weight_norm_interface_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 1, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused__weight_norm_interface_0(in_out_ptr0, in_ptr0, in_ptr1, out_ptr0, xnumel, rnumel, XBLOCK : tl.constexpr):
xnumel = 4
rnumel = 64
RBLOCK: tl.constexpr = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
roffset = 0
rmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
r1 = rindex
x0 = xindex
tmp0 = tl.load(in_ptr0 + (r1 + (64*x0)), xmask, other=0.0)
tmp7 = tl.load(in_ptr1 + (x0), xmask, eviction_policy='evict_last')
tmp1 = tmp0 * tmp0
tmp2 = tl.broadcast_to(tmp1, [XBLOCK, RBLOCK])
tmp4 = tl.where(xmask, tmp2, 0)
tmp5 = tl.sum(tmp4, 1)[:, None]
tmp6 = libdevice.sqrt(tmp5)
tmp8 = tmp7 / tmp6
tmp9 = tmp0 * tmp8
tl.debug_barrier()
tl.store(in_out_ptr0 + (x0), tmp6, xmask)
tl.store(out_ptr0 + (r1 + (64*x0)), tmp9, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/tc/ctcagp37ljugm52zu6ckorigrppqo67voefe2f2odg5r6hyllhyu.py
# Topologically Sorted Source Nodes: [out], Original ATen: [aten.convolution]
# Source node to ATen node mapping:
# out => convolution
# Graph fragment:
# %convolution : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%primals_4, %mul, %primals_3, [1, 1], [0, 0], [1, 1], False, [0, 0], 1), kwargs = {})
triton_poi_fused_convolution_1 = async_compile.triton('triton_poi_fused_convolution_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_1', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_1(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + (x2), tmp2, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4 = args
args.clear()
assert_size_stride(primals_1, (4, 1, 1, 1), (1, 1, 1, 1))
assert_size_stride(primals_2, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_3, (4, ), (1, ))
assert_size_stride(primals_4, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 1, 1, 1), (1, 4, 4, 4), torch.float32)
buf1 = reinterpret_tensor(buf0, (4, 1, 1, 1), (1, 1, 1, 1), 0); del buf0 # reuse
buf2 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [_weight_norm], Original ATen: [aten._weight_norm_interface]
stream0 = get_raw_stream(0)
triton_per_fused__weight_norm_interface_0.run(buf1, primals_2, primals_1, buf2, 4, 64, grid=grid(4), stream=stream0)
# Topologically Sorted Source Nodes: [out], Original ATen: [aten.convolution]
buf3 = extern_kernels.convolution(primals_4, buf2, stride=(1, 1), padding=(0, 0), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf3, (4, 4, 1, 1), (4, 1, 1, 1))
buf4 = buf3; del buf3 # reuse
# Topologically Sorted Source Nodes: [out], Original ATen: [aten.convolution]
triton_poi_fused_convolution_1.run(buf4, primals_3, 16, grid=grid(16), stream=stream0)
del primals_3
return (buf4, buf2, primals_1, primals_2, primals_4, buf1, buf2, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 1, 1, 1), (1, 1, 1, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
class WNConv2d(nn.Module):
def __init__(self, in_channel, out_channel, kernel_size, stride=1,
padding=0, bias=True, activation=None):
super().__init__()
self.conv = nn.utils.weight_norm(nn.Conv2d(in_channel, out_channel,
kernel_size, stride=stride, padding=padding, bias=bias))
self.out_channel = out_channel
if isinstance(kernel_size, int):
kernel_size = [kernel_size, kernel_size]
self.kernel_size = kernel_size
self.activation = activation
def forward(self, input):
out = self.conv(input)
if self.activation is not None:
out = self.activation(out)
return out
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'in_channel': 4, 'out_channel': 4, 'kernel_size': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime.triton_helpers import libdevice
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_per_fused__weight_norm_interface_0(in_out_ptr0, in_ptr0, in_ptr1,
out_ptr0, xnumel, rnumel, XBLOCK: tl.constexpr):
xnumel = 4
RBLOCK: tl.constexpr = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
r1 = rindex
x0 = xindex
tmp0 = tl.load(in_ptr0 + (r1 + 64 * x0), xmask, other=0.0)
tmp7 = tl.load(in_ptr1 + x0, xmask, eviction_policy='evict_last')
tmp1 = tmp0 * tmp0
tmp2 = tl.broadcast_to(tmp1, [XBLOCK, RBLOCK])
tmp4 = tl.where(xmask, tmp2, 0)
tmp5 = tl.sum(tmp4, 1)[:, None]
tmp6 = libdevice.sqrt(tmp5)
tmp8 = tmp7 / tmp6
tmp9 = tmp0 * tmp8
tl.debug_barrier()
tl.store(in_out_ptr0 + x0, tmp6, xmask)
tl.store(out_ptr0 + (r1 + 64 * x0), tmp9, xmask)
@triton.jit
def triton_poi_fused_convolution_1(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl
.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + x2, tmp2, xmask)
def call(args):
primals_1, primals_2, primals_3, primals_4 = args
args.clear()
assert_size_stride(primals_1, (4, 1, 1, 1), (1, 1, 1, 1))
assert_size_stride(primals_2, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_3, (4,), (1,))
assert_size_stride(primals_4, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 1, 1, 1), (1, 4, 4, 4), torch.float32)
buf1 = reinterpret_tensor(buf0, (4, 1, 1, 1), (1, 1, 1, 1), 0)
del buf0
buf2 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
get_raw_stream(0)
triton_per_fused__weight_norm_interface_0[grid(4)](buf1, primals_2,
primals_1, buf2, 4, 64, XBLOCK=1, num_warps=2, num_stages=1)
buf3 = extern_kernels.convolution(primals_4, buf2, stride=(1, 1),
padding=(0, 0), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf3, (4, 4, 1, 1), (4, 1, 1, 1))
buf4 = buf3
del buf3
triton_poi_fused_convolution_1[grid(16)](buf4, primals_3, 16,
XBLOCK=16, num_warps=1, num_stages=1)
del primals_3
return buf4, buf2, primals_1, primals_2, primals_4, buf1, buf2
class WNConv2dNew(nn.Module):
def __init__(self, in_channel, out_channel, kernel_size, stride=1,
padding=0, bias=True, activation=None):
super().__init__()
self.conv = nn.utils.weight_norm(nn.Conv2d(in_channel, out_channel,
kernel_size, stride=stride, padding=padding, bias=bias))
self.out_channel = out_channel
if isinstance(kernel_size, int):
kernel_size = [kernel_size, kernel_size]
self.kernel_size = kernel_size
self.activation = activation
def forward(self, input_0):
primals_3 = self.conv.bias
primals_1 = self.conv.weight_g
primals_2 = self.conv.weight_v
primals_4 = input_0
output = call([primals_1, primals_2, primals_3, primals_4])
return output[0]
| imatge-upc/pixelcoordEDL | WNConv2d | false | 6,873 | [
"MIT"
] | 1 | 353632feed6ac8c93758c1a2a1b7a477e7ff053c | https://github.com/imatge-upc/pixelcoordEDL/tree/353632feed6ac8c93758c1a2a1b7a477e7ff053c | import torch
import torch.nn as nn
class Model(nn.Module):
def __init__(self, in_channel, out_channel, kernel_size, stride=1,
padding=0, bias=True, activation=None):
super().__init__()
self.conv = nn.utils.weight_norm(nn.Conv2d(in_channel, out_channel,
kernel_size, stride=stride, padding=padding, bias=bias))
self.out_channel = out_channel
if isinstance(kernel_size, int):
kernel_size = [kernel_size, kernel_size]
self.kernel_size = kernel_size
self.activation = activation
def forward(self, input):
out = self.conv(input)
if self.activation is not None:
out = self.activation(out)
return out
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4, 4, 4]
|
TransformerDecoderLayer | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/qw/cqw7yoyglmtjad3kirznl5odetqfs3k6pjtnfdbzklyhsdvuvgft.py
# Topologically Sorted Source Nodes: [multi_head_attention_forward], Original ATen: [aten.mul]
# Source node to ATen node mapping:
# multi_head_attention_forward => mul
# Graph fragment:
# %mul : [num_users=2] = call_function[target=torch.ops.aten.mul.Tensor](args = (%permute_3, 1.0), kwargs = {})
triton_poi_fused_mul_0 = async_compile.triton('triton_poi_fused_mul_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_mul_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_mul_0(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = 1.0
tmp4 = tmp2 * tmp3
tl.store(in_out_ptr0 + (x2), tmp4, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/hz/chzi3aam26mikdhljz5x7jlqazm7kpktzeptsf36thgfhsg7ub6a.py
# Topologically Sorted Source Nodes: [multi_head_attention_forward], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# multi_head_attention_forward => amax, exp, sub
# Graph fragment:
# %amax : [num_users=1] = call_function[target=torch.ops.aten.amax.default](args = (%bmm, [-1], True), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%bmm, %amax), kwargs = {})
# %exp : [num_users=2] = call_function[target=torch.ops.aten.exp.default](args = (%sub,), kwargs = {})
triton_poi_fused__softmax_1 = async_compile.triton('triton_poi_fused__softmax_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_1(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp3 = triton_helpers.maximum(tmp1, tmp2)
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp7 = triton_helpers.maximum(tmp5, tmp6)
tmp8 = tmp0 - tmp7
tmp9 = tl_math.exp(tmp8)
tl.store(out_ptr0 + (x2), tmp9, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/em/cem6qbxwbiqnjqybzk5arf2obt5uggy4qs7otwwpovvnrhvdc6h4.py
# Topologically Sorted Source Nodes: [multi_head_attention_forward], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# multi_head_attention_forward => div, sum_1
# Graph fragment:
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%exp, [-1], True), kwargs = {})
# %div : [num_users=2] = call_function[target=torch.ops.aten.div.Tensor](args = (%exp, %sum_1), kwargs = {})
triton_poi_fused__softmax_2 = async_compile.triton('triton_poi_fused__softmax_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_2(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + (x2), tmp8, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/rh/crhjfwyl6xoj5ylcsbbh6lp2vlegits2zkdej3b3wb2q4ddfnejv.py
# Topologically Sorted Source Nodes: [multi_head_attention_forward], Original ATen: [aten.clone]
# Source node to ATen node mapping:
# multi_head_attention_forward => clone
# Graph fragment:
# %clone : [num_users=1] = call_function[target=torch.ops.aten.clone.default](args = (%permute_7,), kwargs = {memory_format: torch.contiguous_format})
triton_poi_fused_clone_3 = async_compile.triton('triton_poi_fused_clone_3', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[4, 4], tile_hint=TileHint.SQUARE,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clone_3', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clone_3(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 4
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x1 = xindex
y0 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + (4*x1)), xmask & ymask)
tl.store(out_ptr0 + (x1 + (4*y0)), tmp0, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/5z/c5zy7julai2lhuinuwjgyl62nx7cyws6ni5poe5jzp7qn532rcgh.py
# Topologically Sorted Source Nodes: [tgt], Original ATen: [aten.add]
# Source node to ATen node mapping:
# tgt => add
# Graph fragment:
# %add : [num_users=2] = call_function[target=torch.ops.aten.add.Tensor](args = (%primals_1, %squeeze), kwargs = {})
triton_poi_fused_add_4 = async_compile.triton('triton_poi_fused_add_4', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_4', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 3, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_4(in_out_ptr0, in_ptr0, in_ptr1, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_out_ptr0 + (x2), xmask)
tmp2 = tl.load(in_ptr1 + (x0), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp4 = tmp0 + tmp3
tl.store(in_out_ptr0 + (x2), tmp4, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/p5/cp5zfsohdolod3ikxheo5zxosgk7sw3fpun4wu4lt34g7uwo2pdt.py
# Topologically Sorted Source Nodes: [leaky_relu], Original ATen: [aten.leaky_relu]
# Source node to ATen node mapping:
# leaky_relu => gt, mul_2, where
# Graph fragment:
# %add_tensor_1 : [num_users=3] = call_function[target=torch.ops.aten.add.Tensor](args = (%mm_default_1, %primals_12), kwargs = {})
# %gt : [num_users=2] = call_function[target=torch.ops.aten.gt.Scalar](args = (%add_tensor_1, 0), kwargs = {})
# %mul_2 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%add_tensor_1, True), kwargs = {})
# %where : [num_users=2] = call_function[target=torch.ops.aten.where.self](args = (%gt, %add_tensor_1, %mul_2), kwargs = {})
triton_poi_fused_leaky_relu_5 = async_compile.triton('triton_poi_fused_leaky_relu_5', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*i1', 3: '*fp32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_leaky_relu_5', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_leaky_relu_5(in_ptr0, in_ptr1, out_ptr0, out_ptr1, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 16
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr1 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = 0.0
tmp4 = tmp2 > tmp3
tmp5 = 1.0
tmp6 = tmp2 * tmp5
tmp7 = tl.where(tmp4, tmp2, tmp6)
tl.store(out_ptr0 + (x2), tmp4, xmask)
tl.store(out_ptr1 + (x2), tmp7, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9, primals_10, primals_11, primals_12, primals_13, primals_14 = args
args.clear()
assert_size_stride(primals_1, (4, 4), (4, 1))
assert_size_stride(primals_2, (12, 4), (4, 1))
assert_size_stride(primals_3, (12, ), (1, ))
assert_size_stride(primals_4, (4, 4), (4, 1))
assert_size_stride(primals_5, (4, ), (1, ))
assert_size_stride(primals_6, (4, 4), (4, 1))
assert_size_stride(primals_7, (12, 4), (4, 1))
assert_size_stride(primals_8, (12, ), (1, ))
assert_size_stride(primals_9, (4, 4), (4, 1))
assert_size_stride(primals_10, (4, ), (1, ))
assert_size_stride(primals_11, (16, 4), (4, 1))
assert_size_stride(primals_12, (16, ), (1, ))
assert_size_stride(primals_13, (4, 16), (16, 1))
assert_size_stride(primals_14, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(primals_1, reinterpret_tensor(primals_2, (4, 4), (1, 4), 0), out=buf0)
buf1 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [multi_head_attention_forward], Original ATen: [aten.addmm]
extern_kernels.addmm(reinterpret_tensor(primals_3, (4, ), (1, ), 4), primals_1, reinterpret_tensor(primals_2, (4, 4), (1, 4), 16), alpha=1, beta=1, out=buf1)
buf2 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [multi_head_attention_forward], Original ATen: [aten.addmm]
extern_kernels.addmm(reinterpret_tensor(primals_3, (4, ), (1, ), 8), primals_1, reinterpret_tensor(primals_2, (4, 4), (1, 4), 32), alpha=1, beta=1, out=buf2)
del primals_2
buf3 = reinterpret_tensor(buf0, (4, 4, 1), (1, 4, 16), 0); del buf0 # reuse
# Topologically Sorted Source Nodes: [multi_head_attention_forward], Original ATen: [aten.mul]
stream0 = get_raw_stream(0)
triton_poi_fused_mul_0.run(buf3, primals_3, 16, grid=grid(16), stream=stream0)
del primals_3
buf4 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [multi_head_attention_forward], Original ATen: [aten.bmm]
extern_kernels.bmm(buf3, reinterpret_tensor(buf1, (4, 1, 4), (1, 1, 4), 0), out=buf4)
buf5 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [multi_head_attention_forward], Original ATen: [aten._softmax]
triton_poi_fused__softmax_1.run(buf4, buf5, 64, grid=grid(64), stream=stream0)
buf6 = buf4; del buf4 # reuse
# Topologically Sorted Source Nodes: [multi_head_attention_forward], Original ATen: [aten._softmax]
triton_poi_fused__softmax_2.run(buf5, buf6, 64, grid=grid(64), stream=stream0)
buf7 = empty_strided_cuda((4, 4, 1), (4, 1, 1), torch.float32)
# Topologically Sorted Source Nodes: [multi_head_attention_forward], Original ATen: [aten.bmm]
extern_kernels.bmm(buf6, reinterpret_tensor(buf2, (4, 4, 1), (1, 4, 1), 0), out=buf7)
buf8 = empty_strided_cuda((4, 4, 1), (4, 1, 1), torch.float32)
# Topologically Sorted Source Nodes: [multi_head_attention_forward], Original ATen: [aten.clone]
triton_poi_fused_clone_3.run(buf7, buf8, 4, 4, grid=grid(4, 4), stream=stream0)
buf9 = reinterpret_tensor(buf7, (4, 4), (4, 1), 0); del buf7 # reuse
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(buf8, (4, 4), (4, 1), 0), reinterpret_tensor(primals_4, (4, 4), (1, 4), 0), out=buf9)
buf10 = buf9; del buf9 # reuse
# Topologically Sorted Source Nodes: [tgt], Original ATen: [aten.add]
triton_poi_fused_add_4.run(buf10, primals_1, primals_5, 16, grid=grid(16), stream=stream0)
del primals_5
buf11 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(buf10, reinterpret_tensor(primals_7, (4, 4), (1, 4), 0), out=buf11)
buf12 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [multi_head_attention_forward_1], Original ATen: [aten.addmm]
extern_kernels.addmm(reinterpret_tensor(primals_8, (4, ), (1, ), 4), primals_6, reinterpret_tensor(primals_7, (4, 4), (1, 4), 16), alpha=1, beta=1, out=buf12)
buf13 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [multi_head_attention_forward_1], Original ATen: [aten.addmm]
extern_kernels.addmm(reinterpret_tensor(primals_8, (4, ), (1, ), 8), primals_6, reinterpret_tensor(primals_7, (4, 4), (1, 4), 32), alpha=1, beta=1, out=buf13)
buf14 = reinterpret_tensor(buf11, (4, 4, 1), (1, 4, 16), 0); del buf11 # reuse
# Topologically Sorted Source Nodes: [multi_head_attention_forward_1], Original ATen: [aten.mul]
triton_poi_fused_mul_0.run(buf14, primals_8, 16, grid=grid(16), stream=stream0)
del primals_8
buf15 = buf5; del buf5 # reuse
# Topologically Sorted Source Nodes: [multi_head_attention_forward_1], Original ATen: [aten.bmm]
extern_kernels.bmm(buf14, reinterpret_tensor(buf12, (4, 1, 4), (1, 1, 4), 0), out=buf15)
buf16 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [multi_head_attention_forward_1], Original ATen: [aten._softmax]
triton_poi_fused__softmax_1.run(buf15, buf16, 64, grid=grid(64), stream=stream0)
buf17 = buf15; del buf15 # reuse
# Topologically Sorted Source Nodes: [multi_head_attention_forward_1], Original ATen: [aten._softmax]
triton_poi_fused__softmax_2.run(buf16, buf17, 64, grid=grid(64), stream=stream0)
buf18 = empty_strided_cuda((4, 4, 1), (4, 1, 1), torch.float32)
# Topologically Sorted Source Nodes: [multi_head_attention_forward_1], Original ATen: [aten.bmm]
extern_kernels.bmm(buf17, reinterpret_tensor(buf13, (4, 4, 1), (1, 4, 1), 0), out=buf18)
buf19 = empty_strided_cuda((4, 4, 1), (4, 1, 1), torch.float32)
# Topologically Sorted Source Nodes: [multi_head_attention_forward_1], Original ATen: [aten.clone]
triton_poi_fused_clone_3.run(buf18, buf19, 4, 4, grid=grid(4, 4), stream=stream0)
buf20 = reinterpret_tensor(buf18, (4, 4), (4, 1), 0); del buf18 # reuse
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(buf19, (4, 4), (4, 1), 0), reinterpret_tensor(primals_9, (4, 4), (1, 4), 0), out=buf20)
buf21 = buf20; del buf20 # reuse
# Topologically Sorted Source Nodes: [tgt_1], Original ATen: [aten.add]
triton_poi_fused_add_4.run(buf21, buf10, primals_10, 16, grid=grid(16), stream=stream0)
del primals_10
buf22 = reinterpret_tensor(buf16, (4, 16), (16, 1), 0); del buf16 # reuse
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(buf21, reinterpret_tensor(primals_11, (4, 16), (1, 4), 0), out=buf22)
buf23 = empty_strided_cuda((4, 16), (16, 1), torch.bool)
buf24 = empty_strided_cuda((4, 16), (16, 1), torch.float32)
# Topologically Sorted Source Nodes: [leaky_relu], Original ATen: [aten.leaky_relu]
triton_poi_fused_leaky_relu_5.run(buf22, primals_12, buf23, buf24, 64, grid=grid(64), stream=stream0)
del buf22
del primals_12
buf25 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(buf24, reinterpret_tensor(primals_13, (16, 4), (1, 16), 0), out=buf25)
buf26 = buf25; del buf25 # reuse
# Topologically Sorted Source Nodes: [tgt_2], Original ATen: [aten.add]
triton_poi_fused_add_4.run(buf26, buf21, primals_14, 16, grid=grid(16), stream=stream0)
del primals_14
return (buf26, primals_1, buf6, reinterpret_tensor(buf8, (4, 4), (4, 1), 0), buf10, primals_6, buf17, reinterpret_tensor(buf19, (4, 4), (4, 1), 0), buf21, buf23, buf24, primals_13, primals_11, primals_9, reinterpret_tensor(buf13, (4, 1, 4), (1, 1, 4), 0), reinterpret_tensor(buf14, (4, 1, 4), (1, 1, 4), 0), reinterpret_tensor(buf12, (4, 4, 1), (1, 4, 1), 0), reinterpret_tensor(primals_7, (4, 4), (4, 1), 0), primals_4, reinterpret_tensor(buf2, (4, 1, 4), (1, 1, 4), 0), reinterpret_tensor(buf3, (4, 1, 4), (1, 1, 4), 0), reinterpret_tensor(buf1, (4, 4, 1), (1, 4, 1), 0), )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((12, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((12, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((12, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_8 = rand_strided((12, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_9 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_10 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_11 = rand_strided((16, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_12 = rand_strided((16, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_13 = rand_strided((4, 16), (16, 1), device='cuda:0', dtype=torch.float32)
primals_14 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9, primals_10, primals_11, primals_12, primals_13, primals_14])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
class TransformerDecoderLayer(nn.Module):
def __init__(self, d_model, nhead, dim_feedforward=16, dropout=0):
super(TransformerDecoderLayer, self).__init__()
self.self_attn = nn.MultiheadAttention(d_model, nhead, dropout=dropout)
self.multihead_attn = nn.MultiheadAttention(d_model, nhead, dropout
=dropout)
self.linear1 = nn.Linear(d_model, dim_feedforward)
self.dropout = nn.Dropout(dropout)
self.linear2 = nn.Linear(dim_feedforward, d_model)
self.dropout1 = nn.Dropout(dropout)
self.dropout2 = nn.Dropout(dropout)
self.dropout3 = nn.Dropout(dropout)
self.activation = nn.LeakyReLU(True)
def forward(self, tgt, memory, tgt_mask=None, memory_mask=None,
tgt_key_padding_mask=None, memory_key_padding_mask=None):
tgt2 = self.self_attn(tgt, tgt, tgt)[0]
tgt = tgt + self.dropout1(tgt2)
tgt2 = self.multihead_attn(tgt, memory, memory)[0]
tgt = tgt + self.dropout2(tgt2)
tgt2 = self.linear2(self.dropout(self.activation(self.linear1(tgt))))
tgt = tgt + self.dropout3(tgt2)
return tgt
def get_inputs():
return [torch.rand([4, 4]), torch.rand([4, 4])]
def get_init_inputs():
return [[], {'d_model': 4, 'nhead': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import math as tl_math
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_mul_0(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = 1.0
tmp4 = tmp2 * tmp3
tl.store(in_out_ptr0 + x2, tmp4, xmask)
@triton.jit
def triton_poi_fused__softmax_1(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp3 = triton_helpers.maximum(tmp1, tmp2)
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp7 = triton_helpers.maximum(tmp5, tmp6)
tmp8 = tmp0 - tmp7
tmp9 = tl_math.exp(tmp8)
tl.store(out_ptr0 + x2, tmp9, xmask)
@triton.jit
def triton_poi_fused__softmax_2(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + x2, tmp8, xmask)
@triton.jit
def triton_poi_fused_clone_3(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK: tl.
constexpr, XBLOCK: tl.constexpr):
ynumel = 4
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x1 = xindex
y0 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + 4 * x1), xmask & ymask)
tl.store(out_ptr0 + (x1 + 4 * y0), tmp0, xmask & ymask)
@triton.jit
def triton_poi_fused_add_4(in_out_ptr0, in_ptr0, in_ptr1, xnumel, XBLOCK:
tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_out_ptr0 + x2, xmask)
tmp2 = tl.load(in_ptr1 + x0, xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp4 = tmp0 + tmp3
tl.store(in_out_ptr0 + x2, tmp4, xmask)
@triton.jit
def triton_poi_fused_leaky_relu_5(in_ptr0, in_ptr1, out_ptr0, out_ptr1,
xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 16
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr1 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = 0.0
tmp4 = tmp2 > tmp3
tmp5 = 1.0
tmp6 = tmp2 * tmp5
tmp7 = tl.where(tmp4, tmp2, tmp6)
tl.store(out_ptr0 + x2, tmp4, xmask)
tl.store(out_ptr1 + x2, tmp7, xmask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7, primals_8, primals_9, primals_10, primals_11, primals_12,
primals_13, primals_14) = args
args.clear()
assert_size_stride(primals_1, (4, 4), (4, 1))
assert_size_stride(primals_2, (12, 4), (4, 1))
assert_size_stride(primals_3, (12,), (1,))
assert_size_stride(primals_4, (4, 4), (4, 1))
assert_size_stride(primals_5, (4,), (1,))
assert_size_stride(primals_6, (4, 4), (4, 1))
assert_size_stride(primals_7, (12, 4), (4, 1))
assert_size_stride(primals_8, (12,), (1,))
assert_size_stride(primals_9, (4, 4), (4, 1))
assert_size_stride(primals_10, (4,), (1,))
assert_size_stride(primals_11, (16, 4), (4, 1))
assert_size_stride(primals_12, (16,), (1,))
assert_size_stride(primals_13, (4, 16), (16, 1))
assert_size_stride(primals_14, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.mm(primals_1, reinterpret_tensor(primals_2, (4, 4),
(1, 4), 0), out=buf0)
buf1 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.addmm(reinterpret_tensor(primals_3, (4,), (1,), 4),
primals_1, reinterpret_tensor(primals_2, (4, 4), (1, 4), 16),
alpha=1, beta=1, out=buf1)
buf2 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.addmm(reinterpret_tensor(primals_3, (4,), (1,), 8),
primals_1, reinterpret_tensor(primals_2, (4, 4), (1, 4), 32),
alpha=1, beta=1, out=buf2)
del primals_2
buf3 = reinterpret_tensor(buf0, (4, 4, 1), (1, 4, 16), 0)
del buf0
get_raw_stream(0)
triton_poi_fused_mul_0[grid(16)](buf3, primals_3, 16, XBLOCK=16,
num_warps=1, num_stages=1)
del primals_3
buf4 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
extern_kernels.bmm(buf3, reinterpret_tensor(buf1, (4, 1, 4), (1, 1,
4), 0), out=buf4)
buf5 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
triton_poi_fused__softmax_1[grid(64)](buf4, buf5, 64, XBLOCK=64,
num_warps=1, num_stages=1)
buf6 = buf4
del buf4
triton_poi_fused__softmax_2[grid(64)](buf5, buf6, 64, XBLOCK=64,
num_warps=1, num_stages=1)
buf7 = empty_strided_cuda((4, 4, 1), (4, 1, 1), torch.float32)
extern_kernels.bmm(buf6, reinterpret_tensor(buf2, (4, 4, 1), (1, 4,
1), 0), out=buf7)
buf8 = empty_strided_cuda((4, 4, 1), (4, 1, 1), torch.float32)
triton_poi_fused_clone_3[grid(4, 4)](buf7, buf8, 4, 4, XBLOCK=4,
YBLOCK=4, num_warps=1, num_stages=1)
buf9 = reinterpret_tensor(buf7, (4, 4), (4, 1), 0)
del buf7
extern_kernels.mm(reinterpret_tensor(buf8, (4, 4), (4, 1), 0),
reinterpret_tensor(primals_4, (4, 4), (1, 4), 0), out=buf9)
buf10 = buf9
del buf9
triton_poi_fused_add_4[grid(16)](buf10, primals_1, primals_5, 16,
XBLOCK=16, num_warps=1, num_stages=1)
del primals_5
buf11 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.mm(buf10, reinterpret_tensor(primals_7, (4, 4), (1,
4), 0), out=buf11)
buf12 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.addmm(reinterpret_tensor(primals_8, (4,), (1,), 4),
primals_6, reinterpret_tensor(primals_7, (4, 4), (1, 4), 16),
alpha=1, beta=1, out=buf12)
buf13 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.addmm(reinterpret_tensor(primals_8, (4,), (1,), 8),
primals_6, reinterpret_tensor(primals_7, (4, 4), (1, 4), 32),
alpha=1, beta=1, out=buf13)
buf14 = reinterpret_tensor(buf11, (4, 4, 1), (1, 4, 16), 0)
del buf11
triton_poi_fused_mul_0[grid(16)](buf14, primals_8, 16, XBLOCK=16,
num_warps=1, num_stages=1)
del primals_8
buf15 = buf5
del buf5
extern_kernels.bmm(buf14, reinterpret_tensor(buf12, (4, 1, 4), (1,
1, 4), 0), out=buf15)
buf16 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
triton_poi_fused__softmax_1[grid(64)](buf15, buf16, 64, XBLOCK=64,
num_warps=1, num_stages=1)
buf17 = buf15
del buf15
triton_poi_fused__softmax_2[grid(64)](buf16, buf17, 64, XBLOCK=64,
num_warps=1, num_stages=1)
buf18 = empty_strided_cuda((4, 4, 1), (4, 1, 1), torch.float32)
extern_kernels.bmm(buf17, reinterpret_tensor(buf13, (4, 4, 1), (1,
4, 1), 0), out=buf18)
buf19 = empty_strided_cuda((4, 4, 1), (4, 1, 1), torch.float32)
triton_poi_fused_clone_3[grid(4, 4)](buf18, buf19, 4, 4, XBLOCK=4,
YBLOCK=4, num_warps=1, num_stages=1)
buf20 = reinterpret_tensor(buf18, (4, 4), (4, 1), 0)
del buf18
extern_kernels.mm(reinterpret_tensor(buf19, (4, 4), (4, 1), 0),
reinterpret_tensor(primals_9, (4, 4), (1, 4), 0), out=buf20)
buf21 = buf20
del buf20
triton_poi_fused_add_4[grid(16)](buf21, buf10, primals_10, 16,
XBLOCK=16, num_warps=1, num_stages=1)
del primals_10
buf22 = reinterpret_tensor(buf16, (4, 16), (16, 1), 0)
del buf16
extern_kernels.mm(buf21, reinterpret_tensor(primals_11, (4, 16), (1,
4), 0), out=buf22)
buf23 = empty_strided_cuda((4, 16), (16, 1), torch.bool)
buf24 = empty_strided_cuda((4, 16), (16, 1), torch.float32)
triton_poi_fused_leaky_relu_5[grid(64)](buf22, primals_12, buf23,
buf24, 64, XBLOCK=64, num_warps=1, num_stages=1)
del buf22
del primals_12
buf25 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.mm(buf24, reinterpret_tensor(primals_13, (16, 4), (1,
16), 0), out=buf25)
buf26 = buf25
del buf25
triton_poi_fused_add_4[grid(16)](buf26, buf21, primals_14, 16,
XBLOCK=16, num_warps=1, num_stages=1)
del primals_14
return (buf26, primals_1, buf6, reinterpret_tensor(buf8, (4, 4), (4, 1),
0), buf10, primals_6, buf17, reinterpret_tensor(buf19, (4, 4), (4,
1), 0), buf21, buf23, buf24, primals_13, primals_11, primals_9,
reinterpret_tensor(buf13, (4, 1, 4), (1, 1, 4), 0),
reinterpret_tensor(buf14, (4, 1, 4), (1, 1, 4), 0),
reinterpret_tensor(buf12, (4, 4, 1), (1, 4, 1), 0),
reinterpret_tensor(primals_7, (4, 4), (4, 1), 0), primals_4,
reinterpret_tensor(buf2, (4, 1, 4), (1, 1, 4), 0),
reinterpret_tensor(buf3, (4, 1, 4), (1, 1, 4), 0),
reinterpret_tensor(buf1, (4, 4, 1), (1, 4, 1), 0))
class TransformerDecoderLayerNew(nn.Module):
def __init__(self, d_model, nhead, dim_feedforward=16, dropout=0):
super(TransformerDecoderLayerNew, self).__init__()
self.self_attn = nn.MultiheadAttention(d_model, nhead, dropout=dropout)
self.multihead_attn = nn.MultiheadAttention(d_model, nhead, dropout
=dropout)
self.linear1 = nn.Linear(d_model, dim_feedforward)
self.dropout = nn.Dropout(dropout)
self.linear2 = nn.Linear(dim_feedforward, d_model)
self.dropout1 = nn.Dropout(dropout)
self.dropout2 = nn.Dropout(dropout)
self.dropout3 = nn.Dropout(dropout)
self.activation = nn.LeakyReLU(True)
def forward(self, input_0, input_1):
primals_2 = self.self_attn.in_proj_weight
primals_3 = self.self_attn.in_proj_bias
primals_1 = self.self_attn.out_proj.weight
primals_5 = self.self_attn.out_proj.bias
primals_7 = self.multihead_attn.in_proj_weight
primals_8 = self.multihead_attn.in_proj_bias
primals_4 = self.multihead_attn.out_proj.weight
primals_10 = self.multihead_attn.out_proj.bias
primals_11 = self.linear1.weight
primals_12 = self.linear1.bias
primals_13 = self.linear2.weight
primals_14 = self.linear2.bias
primals_6 = input_0
primals_9 = input_1
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7, primals_8, primals_9,
primals_10, primals_11, primals_12, primals_13, primals_14])
return output[0]
| imperial-qore/CAROL | TransformerDecoderLayer | false | 6,874 | [
"BSD-3-Clause"
] | 1 | 57dc42c4ddeb9e75eed43a91ceb336a1ecc9c8b9 | https://github.com/imperial-qore/CAROL/tree/57dc42c4ddeb9e75eed43a91ceb336a1ecc9c8b9 | import torch
import torch.nn as nn
class Model(nn.Module):
def __init__(self, d_model, nhead, dim_feedforward=16, dropout=0):
super().__init__()
self.self_attn = nn.MultiheadAttention(d_model, nhead, dropout=dropout)
self.multihead_attn = nn.MultiheadAttention(d_model, nhead, dropout
=dropout)
self.linear1 = nn.Linear(d_model, dim_feedforward)
self.dropout = nn.Dropout(dropout)
self.linear2 = nn.Linear(dim_feedforward, d_model)
self.dropout1 = nn.Dropout(dropout)
self.dropout2 = nn.Dropout(dropout)
self.dropout3 = nn.Dropout(dropout)
self.activation = nn.LeakyReLU(True)
def forward(self, tgt, memory, tgt_mask=None, memory_mask=None,
tgt_key_padding_mask=None, memory_key_padding_mask=None):
tgt2 = self.self_attn(tgt, tgt, tgt)[0]
tgt = tgt + self.dropout1(tgt2)
tgt2 = self.multihead_attn(tgt, memory, memory)[0]
tgt = tgt + self.dropout2(tgt2)
tgt2 = self.linear2(self.dropout(self.activation(self.linear1(tgt))))
tgt = tgt + self.dropout3(tgt2)
return tgt
def get_inputs():
return [torch.rand([4, 4]), torch.rand([4, 4])]
def get_init_inputs():
return [4, 4]
|
FocalLossV1 | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/3p/c3pzst4ut3m4lzweova7e7vpyokobuvqch7kkb2vnvmj6tlxss4a.py
# Topologically Sorted Source Nodes: [alpha, setitem, eq_1, probs, sub, pt, sub_1, pow_1, mul, ce_loss, loss, loss_1], Original ATen: [aten.fill, aten.lift_fresh, aten.index_put, aten.eq, aten.sigmoid, aten.rsub, aten.where, aten.pow, aten.mul, aten.binary_cross_entropy_with_logits, aten.mean]
# Source node to ATen node mapping:
# alpha => full_default
# ce_loss => abs_1, exp, full_default_2, log1p, minimum, mul, neg, sub_1, sub_2, sub_3
# eq_1 => eq_1
# loss => mul_2
# loss_1 => mean
# mul => mul_1
# pow_1 => pow_1
# probs => sigmoid
# pt => where
# setitem => full_default_1, index_put
# sub => sub
# sub_1 => sub_4
# Graph fragment:
# %full_default : [num_users=1] = call_function[target=torch.ops.aten.full.default](args = ([4, 4, 4, 4], 0.75), kwargs = {dtype: torch.float32, layout: torch.strided, device: cuda:0, pin_memory: False})
# %full_default_1 : [num_users=1] = call_function[target=torch.ops.aten.full.default](args = ([], 0.25), kwargs = {dtype: torch.float32, layout: torch.strided, device: cpu, pin_memory: False})
# %index_put : [num_users=1] = call_function[target=torch.ops.aten.index_put_.default](args = (%full_default, [%eq], %full_default_1), kwargs = {})
# %eq_1 : [num_users=1] = call_function[target=torch.ops.aten.eq.Scalar](args = (%arg1_1, 1), kwargs = {})
# %sigmoid : [num_users=2] = call_function[target=torch.ops.aten.sigmoid.default](args = (%arg0_1,), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (1, %sigmoid), kwargs = {})
# %where : [num_users=1] = call_function[target=torch.ops.aten.where.self](args = (%eq_1, %sigmoid, %sub), kwargs = {})
# %sub_4 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (1, %where), kwargs = {})
# %pow_1 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%sub_4, 2), kwargs = {})
# %mul_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%index_put, %pow_1), kwargs = {})
# %sub_1 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (1, %arg1_1), kwargs = {})
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sub_1, %arg0_1), kwargs = {})
# %full_default_2 : [num_users=1] = call_function[target=torch.ops.aten.full.default](args = ([], 0), kwargs = {dtype: torch.float32, layout: torch.strided, device: cuda:0, pin_memory: False})
# %minimum : [num_users=1] = call_function[target=torch.ops.aten.minimum.default](args = (%full_default_2, %arg0_1), kwargs = {})
# %abs_1 : [num_users=1] = call_function[target=torch.ops.aten.abs.default](args = (%arg0_1,), kwargs = {})
# %neg : [num_users=1] = call_function[target=torch.ops.aten.neg.default](args = (%abs_1,), kwargs = {})
# %exp : [num_users=1] = call_function[target=torch.ops.aten.exp.default](args = (%neg,), kwargs = {})
# %log1p : [num_users=1] = call_function[target=torch.ops.aten.log1p.default](args = (%exp,), kwargs = {})
# %sub_2 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%minimum, %log1p), kwargs = {})
# %sub_3 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%mul, %sub_2), kwargs = {})
# %mul_2 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%mul_1, %sub_3), kwargs = {})
# %mean : [num_users=1] = call_function[target=torch.ops.aten.mean.default](args = (%mul_2,), kwargs = {})
triton_per_fused_binary_cross_entropy_with_logits_eq_fill_index_put_lift_fresh_mean_mul_pow_rsub_sigmoid_where_0 = async_compile.triton('triton_per_fused_binary_cross_entropy_with_logits_eq_fill_index_put_lift_fresh_mean_mul_pow_rsub_sigmoid_where_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[1, 256],
reduction_hint=ReductionHint.INNER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {3: 1}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 4), equal_to_1=(3,))]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused_binary_cross_entropy_with_logits_eq_fill_index_put_lift_fresh_mean_mul_pow_rsub_sigmoid_where_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': True, 'num_load': 2, 'num_reduction': 1, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused_binary_cross_entropy_with_logits_eq_fill_index_put_lift_fresh_mean_mul_pow_rsub_sigmoid_where_0(in_out_ptr0, in_ptr0, in_ptr1, xnumel, rnumel):
xnumel = 1
XBLOCK: tl.constexpr = 1
rnumel = 256
RBLOCK: tl.constexpr = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = tl.full([1], xoffset, tl.int32)
xmask = tl.full([RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[:]
roffset = 0
rmask = tl.full([RBLOCK], True, tl.int1)
r0 = rindex
tmp0 = tl.load(in_ptr0 + (r0), None)
tmp6 = tl.load(in_ptr1 + (r0), None)
tmp1 = 1.0
tmp2 = tmp0 == tmp1
tmp3 = 0.25
tmp4 = 0.75
tmp5 = tl.where(tmp2, tmp3, tmp4)
tmp7 = tl.sigmoid(tmp6)
tmp8 = tmp1 - tmp7
tmp9 = tl.where(tmp2, tmp7, tmp8)
tmp10 = tmp1 - tmp9
tmp11 = tmp10 * tmp10
tmp12 = tmp5 * tmp11
tmp13 = tmp1 - tmp0
tmp14 = tmp13 * tmp6
tmp15 = 0.0
tmp16 = triton_helpers.minimum(tmp15, tmp6)
tmp17 = tl_math.abs(tmp6)
tmp18 = -tmp17
tmp19 = tl_math.exp(tmp18)
tmp20 = libdevice.log1p(tmp19)
tmp21 = tmp16 - tmp20
tmp22 = tmp14 - tmp21
tmp23 = tmp12 * tmp22
tmp24 = tl.broadcast_to(tmp23, [RBLOCK])
tmp26 = triton_helpers.promote_to_tensor(tl.sum(tmp24, 0))
tmp27 = 256.0
tmp28 = tmp26 / tmp27
tl.debug_barrier()
tl.store(in_out_ptr0 + (tl.full([1], 0, tl.int32)), tmp28, None)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, arg1_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg1_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf1 = empty_strided_cuda((), (), torch.float32)
buf2 = buf1; del buf1 # reuse
# Topologically Sorted Source Nodes: [alpha, setitem, eq_1, probs, sub, pt, sub_1, pow_1, mul, ce_loss, loss, loss_1], Original ATen: [aten.fill, aten.lift_fresh, aten.index_put, aten.eq, aten.sigmoid, aten.rsub, aten.where, aten.pow, aten.mul, aten.binary_cross_entropy_with_logits, aten.mean]
stream0 = get_raw_stream(0)
triton_per_fused_binary_cross_entropy_with_logits_eq_fill_index_put_lift_fresh_mean_mul_pow_rsub_sigmoid_where_0.run(buf2, arg1_1, arg0_1, 1, 256, grid=grid(1), stream=stream0)
del arg0_1
del arg1_1
return (buf2, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
arg1_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([arg0_1, arg1_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
class FocalLossV1(nn.Module):
def __init__(self, alpha=0.25, gamma=2, reduction='mean'):
super(FocalLossV1, self).__init__()
self.alpha = alpha
self.gamma = gamma
self.reduction = reduction
self.crit = nn.BCEWithLogitsLoss(reduction='none')
def forward(self, logits, label):
"""
logits and label have same shape, and label data type is long
args:
logits: tensor of shape (N, ...)
label: tensor of shape(N, ...)
"""
logits = logits.float()
with torch.no_grad():
alpha = torch.empty_like(logits).fill_(1 - self.alpha)
alpha[label == 1] = self.alpha
probs = torch.sigmoid(logits)
pt = torch.where(label == 1, probs, 1 - probs)
ce_loss = self.crit(logits, label.float())
loss = alpha * torch.pow(1 - pt, self.gamma) * ce_loss
if self.reduction == 'mean':
loss = loss.mean()
if self.reduction == 'sum':
loss = loss.sum()
return loss
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {}]
| import torch
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_per_fused_binary_cross_entropy_with_logits_eq_fill_index_put_lift_fresh_mean_mul_pow_rsub_sigmoid_where_0(
in_out_ptr0, in_ptr0, in_ptr1, xnumel, rnumel):
XBLOCK: tl.constexpr = 1
RBLOCK: tl.constexpr = 256
xoffset = tl.program_id(0) * XBLOCK
tl.full([1], xoffset, tl.int32)
tl.full([RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[:]
tl.full([RBLOCK], True, tl.int1)
r0 = rindex
tmp0 = tl.load(in_ptr0 + r0, None)
tmp6 = tl.load(in_ptr1 + r0, None)
tmp1 = 1.0
tmp2 = tmp0 == tmp1
tmp3 = 0.25
tmp4 = 0.75
tmp5 = tl.where(tmp2, tmp3, tmp4)
tmp7 = tl.sigmoid(tmp6)
tmp8 = tmp1 - tmp7
tmp9 = tl.where(tmp2, tmp7, tmp8)
tmp10 = tmp1 - tmp9
tmp11 = tmp10 * tmp10
tmp12 = tmp5 * tmp11
tmp13 = tmp1 - tmp0
tmp14 = tmp13 * tmp6
tmp15 = 0.0
tmp16 = triton_helpers.minimum(tmp15, tmp6)
tmp17 = tl_math.abs(tmp6)
tmp18 = -tmp17
tmp19 = tl_math.exp(tmp18)
tmp20 = libdevice.log1p(tmp19)
tmp21 = tmp16 - tmp20
tmp22 = tmp14 - tmp21
tmp23 = tmp12 * tmp22
tmp24 = tl.broadcast_to(tmp23, [RBLOCK])
tmp26 = triton_helpers.promote_to_tensor(tl.sum(tmp24, 0))
tmp27 = 256.0
tmp28 = tmp26 / tmp27
tl.debug_barrier()
tl.store(in_out_ptr0 + tl.full([1], 0, tl.int32), tmp28, None)
def call(args):
arg0_1, arg1_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg1_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf1 = empty_strided_cuda((), (), torch.float32)
buf2 = buf1
del buf1
get_raw_stream(0)
triton_per_fused_binary_cross_entropy_with_logits_eq_fill_index_put_lift_fresh_mean_mul_pow_rsub_sigmoid_where_0[
grid(1)](buf2, arg1_1, arg0_1, 1, 256, num_warps=2, num_stages=1)
del arg0_1
del arg1_1
return buf2,
class FocalLossV1New(nn.Module):
def __init__(self, alpha=0.25, gamma=2, reduction='mean'):
super(FocalLossV1New, self).__init__()
self.alpha = alpha
self.gamma = gamma
self.reduction = reduction
self.crit = nn.BCEWithLogitsLoss(reduction='none')
def forward(self, input_0, input_1):
arg0_1 = input_0
arg1_1 = input_1
output = call([arg0_1, arg1_1])
return output[0]
| imvladikon/pytorch-loss | FocalLossV1 | false | 6,875 | [
"MIT"
] | 1 | 6cfaabe1be898e1ff000b3dffb46d0ef09096f6b | https://github.com/imvladikon/pytorch-loss/tree/6cfaabe1be898e1ff000b3dffb46d0ef09096f6b | import torch
import torch.nn as nn
class Model(nn.Module):
def __init__(self, alpha=0.25, gamma=2, reduction='mean'):
super().__init__()
self.alpha = alpha
self.gamma = gamma
self.reduction = reduction
self.crit = nn.BCEWithLogitsLoss(reduction='none')
def forward(self, logits, label):
"""
logits and label have same shape, and label data type is long
args:
logits: tensor of shape (N, ...)
label: tensor of shape(N, ...)
"""
logits = logits.float()
with torch.no_grad():
alpha = torch.empty_like(logits).fill_(1 - self.alpha)
alpha[label == 1] = self.alpha
probs = torch.sigmoid(logits)
pt = torch.where(label == 1, probs, 1 - probs)
ce_loss = self.crit(logits, label.float())
loss = alpha * torch.pow(1 - pt, self.gamma) * ce_loss
if self.reduction == 'mean':
loss = loss.mean()
if self.reduction == 'sum':
loss = loss.sum()
return loss
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return []
|
SoftDiceLossV1 | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/vc/cvcfr6g466rjwpqyuoei5b3fjc2q6re5nflgvwlz5ocdklbhptjy.py
# Topologically Sorted Source Nodes: [probs, mul, numer, add, denor], Original ATen: [aten.sigmoid, aten.mul, aten.sum, aten.add]
# Source node to ATen node mapping:
# add => add
# denor => sum_2
# mul => mul
# numer => sum_1
# probs => sigmoid
# Graph fragment:
# %sigmoid : [num_users=2] = call_function[target=torch.ops.aten.sigmoid.default](args = (%arg0_1,), kwargs = {})
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sigmoid, %arg1_1), kwargs = {})
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%mul, [1, 2]), kwargs = {})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%sigmoid, %arg1_1), kwargs = {})
# %sum_2 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%add, [1, 2]), kwargs = {})
triton_per_fused_add_mul_sigmoid_sum_0 = async_compile.triton('triton_per_fused_add_mul_sigmoid_sum_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[16, 16],
reduction_hint=ReductionHint.DEFAULT,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32', 5: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused_add_mul_sigmoid_sum_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 2, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused_add_mul_sigmoid_sum_0(in_ptr0, in_ptr1, out_ptr0, out_ptr1, xnumel, rnumel, XBLOCK : tl.constexpr):
xnumel = 16
rnumel = 16
RBLOCK: tl.constexpr = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
roffset = 0
rmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
r2 = rindex
x0 = xindex % 4
x1 = (xindex // 4)
x3 = xindex
tmp0 = tl.load(in_ptr0 + (x0 + (4*r2) + (64*x1)), xmask, other=0.0)
tmp2 = tl.load(in_ptr1 + (x0 + (4*r2) + (64*x1)), xmask, other=0.0)
tmp1 = tl.sigmoid(tmp0)
tmp3 = tmp1 * tmp2
tmp4 = tl.broadcast_to(tmp3, [XBLOCK, RBLOCK])
tmp6 = tl.where(xmask, tmp4, 0)
tmp7 = tl.sum(tmp6, 1)[:, None]
tmp8 = tmp1 + tmp2
tmp9 = tl.broadcast_to(tmp8, [XBLOCK, RBLOCK])
tmp11 = tl.where(xmask, tmp9, 0)
tmp12 = tl.sum(tmp11, 1)[:, None]
tl.store(out_ptr0 + (x3), tmp7, xmask)
tl.store(out_ptr1 + (x3), tmp12, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/2x/c2xcxojd4xdl4urjqd4gxqrmsugot4ql5kir7dkhanipimsb3jvz.py
# Topologically Sorted Source Nodes: [mul_1, add_1, add_2, truediv, loss, loss_1], Original ATen: [aten.mul, aten.add, aten.div, aten.rsub, aten.mean]
# Source node to ATen node mapping:
# add_1 => add_1
# add_2 => add_2
# loss => sub
# loss_1 => mean
# mul_1 => mul_1
# truediv => div
# Graph fragment:
# %mul_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sum_1, 2), kwargs = {})
# %add_1 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul_1, 1), kwargs = {})
# %add_2 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%sum_2, 1), kwargs = {})
# %div : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%add_1, %add_2), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (1.0, %div), kwargs = {})
# %mean : [num_users=1] = call_function[target=torch.ops.aten.mean.default](args = (%sub,), kwargs = {})
triton_per_fused_add_div_mean_mul_rsub_1 = async_compile.triton('triton_per_fused_add_div_mean_mul_rsub_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[1, 16],
reduction_hint=ReductionHint.INNER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {3: 1}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 4), equal_to_1=(3,))]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused_add_div_mean_mul_rsub_1', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 1, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused_add_div_mean_mul_rsub_1(in_out_ptr0, in_ptr0, in_ptr1, xnumel, rnumel, XBLOCK : tl.constexpr):
xnumel = 1
rnumel = 16
RBLOCK: tl.constexpr = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[None, :]
roffset = 0
rmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
r0 = rindex
tmp0 = tl.load(in_ptr0 + (r0), None)
tmp5 = tl.load(in_ptr1 + (r0), None)
tmp1 = 2.0
tmp2 = tmp0 * tmp1
tmp3 = 1.0
tmp4 = tmp2 + tmp3
tmp6 = tmp5 + tmp3
tmp7 = tmp4 / tmp6
tmp8 = tmp3 - tmp7
tmp9 = tl.broadcast_to(tmp8, [XBLOCK, RBLOCK])
tmp11 = tl.sum(tmp9, 1)[:, None]
tmp12 = 16.0
tmp13 = tmp11 / tmp12
tl.debug_barrier()
tl.store(in_out_ptr0 + (tl.full([XBLOCK, 1], 0, tl.int32)), tmp13, None)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, arg1_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg1_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
buf1 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [probs, mul, numer, add, denor], Original ATen: [aten.sigmoid, aten.mul, aten.sum, aten.add]
stream0 = get_raw_stream(0)
triton_per_fused_add_mul_sigmoid_sum_0.run(arg0_1, arg1_1, buf0, buf1, 16, 16, grid=grid(16), stream=stream0)
del arg0_1
del arg1_1
buf2 = empty_strided_cuda((), (), torch.float32)
buf3 = buf2; del buf2 # reuse
# Topologically Sorted Source Nodes: [mul_1, add_1, add_2, truediv, loss, loss_1], Original ATen: [aten.mul, aten.add, aten.div, aten.rsub, aten.mean]
triton_per_fused_add_div_mean_mul_rsub_1.run(buf3, buf0, buf1, 1, 16, grid=grid(1), stream=stream0)
del buf0
del buf1
return (buf3, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
arg1_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([arg0_1, arg1_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
class SoftDiceLossV1(nn.Module):
"""
soft-dice loss, useful in binary segmentation
"""
def __init__(self, p=1, smooth=1, reduction='mean'):
super(SoftDiceLossV1, self).__init__()
self.p = p
self.smooth = smooth
self.reduction = reduction
def forward(self, logits, labels):
"""
args: logits: tensor of shape (N, H, W)
args: label: tensor of shape(N, H, W)
"""
probs = torch.sigmoid(logits)
numer = (probs * labels).sum(dim=(1, 2))
denor = (probs.pow(self.p) + labels).sum(dim=(1, 2))
loss = 1.0 - (2 * numer + self.smooth) / (denor + self.smooth)
if self.reduction == 'mean':
loss = loss.mean()
elif self.reduction == 'sum':
loss = loss.sum()
return loss
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {}]
| import torch
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_per_fused_add_mul_sigmoid_sum_0(in_ptr0, in_ptr1, out_ptr0,
out_ptr1, xnumel, rnumel, XBLOCK: tl.constexpr):
xnumel = 16
RBLOCK: tl.constexpr = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
r2 = rindex
x0 = xindex % 4
x1 = xindex // 4
x3 = xindex
tmp0 = tl.load(in_ptr0 + (x0 + 4 * r2 + 64 * x1), xmask, other=0.0)
tmp2 = tl.load(in_ptr1 + (x0 + 4 * r2 + 64 * x1), xmask, other=0.0)
tmp1 = tl.sigmoid(tmp0)
tmp3 = tmp1 * tmp2
tmp4 = tl.broadcast_to(tmp3, [XBLOCK, RBLOCK])
tmp6 = tl.where(xmask, tmp4, 0)
tmp7 = tl.sum(tmp6, 1)[:, None]
tmp8 = tmp1 + tmp2
tmp9 = tl.broadcast_to(tmp8, [XBLOCK, RBLOCK])
tmp11 = tl.where(xmask, tmp9, 0)
tmp12 = tl.sum(tmp11, 1)[:, None]
tl.store(out_ptr0 + x3, tmp7, xmask)
tl.store(out_ptr1 + x3, tmp12, xmask)
@triton.jit
def triton_per_fused_add_div_mean_mul_rsub_1(in_out_ptr0, in_ptr0, in_ptr1,
xnumel, rnumel, XBLOCK: tl.constexpr):
RBLOCK: tl.constexpr = 16
xoffset = tl.program_id(0) * XBLOCK
xoffset + tl.arange(0, XBLOCK)[:, None]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[None, :]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
r0 = rindex
tmp0 = tl.load(in_ptr0 + r0, None)
tmp5 = tl.load(in_ptr1 + r0, None)
tmp1 = 2.0
tmp2 = tmp0 * tmp1
tmp3 = 1.0
tmp4 = tmp2 + tmp3
tmp6 = tmp5 + tmp3
tmp7 = tmp4 / tmp6
tmp8 = tmp3 - tmp7
tmp9 = tl.broadcast_to(tmp8, [XBLOCK, RBLOCK])
tmp11 = tl.sum(tmp9, 1)[:, None]
tmp12 = 16.0
tmp13 = tmp11 / tmp12
tl.debug_barrier()
tl.store(in_out_ptr0 + tl.full([XBLOCK, 1], 0, tl.int32), tmp13, None)
def call(args):
arg0_1, arg1_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg1_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
buf1 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
get_raw_stream(0)
triton_per_fused_add_mul_sigmoid_sum_0[grid(16)](arg0_1, arg1_1,
buf0, buf1, 16, 16, XBLOCK=1, num_warps=2, num_stages=1)
del arg0_1
del arg1_1
buf2 = empty_strided_cuda((), (), torch.float32)
buf3 = buf2
del buf2
triton_per_fused_add_div_mean_mul_rsub_1[grid(1)](buf3, buf0, buf1,
1, 16, XBLOCK=1, num_warps=2, num_stages=1)
del buf0
del buf1
return buf3,
class SoftDiceLossV1New(nn.Module):
"""
soft-dice loss, useful in binary segmentation
"""
def __init__(self, p=1, smooth=1, reduction='mean'):
super(SoftDiceLossV1New, self).__init__()
self.p = p
self.smooth = smooth
self.reduction = reduction
def forward(self, input_0, input_1):
arg0_1 = input_0
arg1_1 = input_1
output = call([arg0_1, arg1_1])
return output[0]
| imvladikon/pytorch-loss | SoftDiceLossV1 | false | 6,876 | [
"MIT"
] | 1 | 6cfaabe1be898e1ff000b3dffb46d0ef09096f6b | https://github.com/imvladikon/pytorch-loss/tree/6cfaabe1be898e1ff000b3dffb46d0ef09096f6b | import torch
import torch.nn as nn
class Model(nn.Module):
"""
soft-dice loss, useful in binary segmentation
"""
def __init__(self, p=1, smooth=1, reduction='mean'):
super().__init__()
self.p = p
self.smooth = smooth
self.reduction = reduction
def forward(self, logits, labels):
"""
args: logits: tensor of shape (N, H, W)
args: label: tensor of shape(N, H, W)
"""
probs = torch.sigmoid(logits)
numer = (probs * labels).sum(dim=(1, 2))
denor = (probs.pow(self.p) + labels).sum(dim=(1, 2))
loss = 1.0 - (2 * numer + self.smooth) / (denor + self.smooth)
if self.reduction == 'mean':
loss = loss.mean()
elif self.reduction == 'sum':
loss = loss.sum()
return loss
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return []
|
CoordConv2d | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/du/cduoasid3d53iag3ldazcyzvaearrvnqdqh6extmgwnl37hpllzm.py
# Topologically Sorted Source Nodes: [feat], Original ATen: [aten.cat]
# Source node to ATen node mapping:
# feat => cat
# Graph fragment:
# %cat : [num_users=2] = call_function[target=torch.ops.aten.cat.default](args = ([%expand_2, %expand_3, %primals_1], 1), kwargs = {})
triton_poi_fused_cat_0 = async_compile.triton('triton_poi_fused_cat_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[512],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_cat_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_cat_0(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 384
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = (xindex // 16) % 6
x1 = (xindex // 4) % 4
x0 = xindex % 4
x3 = (xindex // 96)
x6 = xindex % 16
x7 = xindex
tmp0 = x2
tmp1 = tl.full([1], 0, tl.int64)
tmp2 = tmp0 >= tmp1
tmp3 = tl.full([1], 1, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = x1
tmp6 = tmp5.to(tl.float32)
tmp7 = 2.0
tmp8 = tmp6 < tmp7
tmp9 = 0.6666666666666666
tmp10 = tmp6 * tmp9
tmp11 = -1.0
tmp12 = tmp10 + tmp11
tmp13 = 3 + ((-1)*x1)
tmp14 = tmp13.to(tl.float32)
tmp15 = tmp14 * tmp9
tmp16 = 1.0
tmp17 = tmp16 - tmp15
tmp18 = tl.where(tmp8, tmp12, tmp17)
tmp19 = tl.full(tmp18.shape, 0.0, tmp18.dtype)
tmp20 = tl.where(tmp4, tmp18, tmp19)
tmp21 = tmp0 >= tmp3
tmp22 = tl.full([1], 2, tl.int64)
tmp23 = tmp0 < tmp22
tmp24 = tmp21 & tmp23
tmp25 = x0
tmp26 = tmp25.to(tl.float32)
tmp27 = tmp26 < tmp7
tmp28 = tmp26 * tmp9
tmp29 = tmp28 + tmp11
tmp30 = 3 + ((-1)*x0)
tmp31 = tmp30.to(tl.float32)
tmp32 = tmp31 * tmp9
tmp33 = tmp16 - tmp32
tmp34 = tl.where(tmp27, tmp29, tmp33)
tmp35 = tl.full(tmp34.shape, 0.0, tmp34.dtype)
tmp36 = tl.where(tmp24, tmp34, tmp35)
tmp37 = tmp0 >= tmp22
tmp38 = tl.full([1], 6, tl.int64)
tmp39 = tmp0 < tmp38
tmp40 = tl.load(in_ptr0 + (x6 + (16*((-2) + x2)) + (64*x3)), tmp37 & xmask, other=0.0)
tmp41 = tl.where(tmp24, tmp36, tmp40)
tmp42 = tl.where(tmp4, tmp20, tmp41)
tl.store(out_ptr0 + (x7), tmp42, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/32/c32v7egt4mupqssam3gmac2qgv3ujprjybthsgweflmot256qqw7.py
# Topologically Sorted Source Nodes: [conv2d], Original ATen: [aten.convolution]
# Source node to ATen node mapping:
# conv2d => convolution
# Graph fragment:
# %convolution : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%cat, %primals_2, %primals_3, [1, 1], [1, 1], [1, 1], False, [0, 0], 1), kwargs = {})
triton_poi_fused_convolution_1 = async_compile.triton('triton_poi_fused_convolution_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_1', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_1(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = (xindex // 16) % 4
tmp0 = tl.load(in_out_ptr0 + (x3), xmask)
tmp1 = tl.load(in_ptr0 + (x1), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + (x3), tmp2, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (4, 6, 3, 3), (54, 9, 3, 1))
assert_size_stride(primals_3, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 6, 4, 4), (96, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [feat], Original ATen: [aten.cat]
stream0 = get_raw_stream(0)
triton_poi_fused_cat_0.run(primals_1, buf0, 384, grid=grid(384), stream=stream0)
del primals_1
# Topologically Sorted Source Nodes: [conv2d], Original ATen: [aten.convolution]
buf1 = extern_kernels.convolution(buf0, primals_2, stride=(1, 1), padding=(1, 1), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf1, (4, 4, 4, 4), (64, 16, 4, 1))
buf2 = buf1; del buf1 # reuse
# Topologically Sorted Source Nodes: [conv2d], Original ATen: [aten.convolution]
triton_poi_fused_convolution_1.run(buf2, primals_3, 256, grid=grid(256), stream=stream0)
del primals_3
return (buf2, primals_2, buf0, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 6, 3, 3), (54, 9, 3, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.nn.functional as F
class CoordConv2d(nn.Conv2d):
def __init__(self, in_chan, out_chan, kernel_size=3, stride=1, padding=
1, dilation=1, groups=1, bias=True):
super(CoordConv2d, self).__init__(in_chan + 2, out_chan,
kernel_size, stride=stride, padding=padding, dilation=dilation,
groups=groups, bias=bias)
def forward(self, x):
batchsize, H, W = x.size(0), x.size(2), x.size(3)
h_range = torch.linspace(-1, 1, H, device=x.device, dtype=x.dtype)
w_range = torch.linspace(-1, 1, W, device=x.device, dtype=x.dtype)
h_chan, w_chan = torch.meshgrid(h_range, w_range)
h_chan = h_chan.expand([batchsize, 1, -1, -1])
w_chan = w_chan.expand([batchsize, 1, -1, -1])
feat = torch.cat([h_chan, w_chan, x], dim=1)
return F.conv2d(feat, self.weight, self.bias, self.stride, self.
padding, self.dilation, self.groups)
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'in_chan': 4, 'out_chan': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_poi_fused_cat_0(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 384
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex // 16 % 6
x1 = xindex // 4 % 4
x0 = xindex % 4
x3 = xindex // 96
x6 = xindex % 16
x7 = xindex
tmp0 = x2
tl.full([1], 0, tl.int64)
tmp3 = tl.full([1], 1, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = x1
tmp6 = tmp5.to(tl.float32)
tmp7 = 2.0
tmp8 = tmp6 < tmp7
tmp9 = 0.6666666666666666
tmp10 = tmp6 * tmp9
tmp11 = -1.0
tmp12 = tmp10 + tmp11
tmp13 = 3 + -1 * x1
tmp14 = tmp13.to(tl.float32)
tmp15 = tmp14 * tmp9
tmp16 = 1.0
tmp17 = tmp16 - tmp15
tmp18 = tl.where(tmp8, tmp12, tmp17)
tmp19 = tl.full(tmp18.shape, 0.0, tmp18.dtype)
tmp20 = tl.where(tmp4, tmp18, tmp19)
tmp21 = tmp0 >= tmp3
tmp22 = tl.full([1], 2, tl.int64)
tmp23 = tmp0 < tmp22
tmp24 = tmp21 & tmp23
tmp25 = x0
tmp26 = tmp25.to(tl.float32)
tmp27 = tmp26 < tmp7
tmp28 = tmp26 * tmp9
tmp29 = tmp28 + tmp11
tmp30 = 3 + -1 * x0
tmp31 = tmp30.to(tl.float32)
tmp32 = tmp31 * tmp9
tmp33 = tmp16 - tmp32
tmp34 = tl.where(tmp27, tmp29, tmp33)
tmp35 = tl.full(tmp34.shape, 0.0, tmp34.dtype)
tmp36 = tl.where(tmp24, tmp34, tmp35)
tmp37 = tmp0 >= tmp22
tl.full([1], 6, tl.int64)
tmp40 = tl.load(in_ptr0 + (x6 + 16 * (-2 + x2) + 64 * x3), tmp37 &
xmask, other=0.0)
tmp41 = tl.where(tmp24, tmp36, tmp40)
tmp42 = tl.where(tmp4, tmp20, tmp41)
tl.store(out_ptr0 + x7, tmp42, xmask)
@triton.jit
def triton_poi_fused_convolution_1(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl
.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = xindex // 16 % 4
tmp0 = tl.load(in_out_ptr0 + x3, xmask)
tmp1 = tl.load(in_ptr0 + x1, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + x3, tmp2, xmask)
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (4, 6, 3, 3), (54, 9, 3, 1))
assert_size_stride(primals_3, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 6, 4, 4), (96, 16, 4, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_cat_0[grid(384)](primals_1, buf0, 384, XBLOCK=128,
num_warps=4, num_stages=1)
del primals_1
buf1 = extern_kernels.convolution(buf0, primals_2, stride=(1, 1),
padding=(1, 1), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf1, (4, 4, 4, 4), (64, 16, 4, 1))
buf2 = buf1
del buf1
triton_poi_fused_convolution_1[grid(256)](buf2, primals_3, 256,
XBLOCK=256, num_warps=4, num_stages=1)
del primals_3
return buf2, primals_2, buf0
class CoordConv2dNew(nn.Conv2d):
def __init__(self, in_chan, out_chan, kernel_size=3, stride=1, padding=
1, dilation=1, groups=1, bias=True):
super(CoordConv2dNew, self).__init__(in_chan + 2, out_chan,
kernel_size, stride=stride, padding=padding, dilation=dilation,
groups=groups, bias=bias)
def forward(self, input_0):
primals_2 = self.weight
primals_3 = self.bias
primals_1 = input_0
output = call([primals_1, primals_2, primals_3])
return output[0]
| imvladikon/pytorch-loss | CoordConv2d | false | 6,877 | [
"MIT"
] | 1 | 6cfaabe1be898e1ff000b3dffb46d0ef09096f6b | https://github.com/imvladikon/pytorch-loss/tree/6cfaabe1be898e1ff000b3dffb46d0ef09096f6b | import torch
import torch.nn as nn
import torch.nn.functional as F
class Model(nn.Conv2d):
def __init__(self, in_chan, out_chan, kernel_size=3, stride=1, padding=
1, dilation=1, groups=1, bias=True):
super().__init__(in_chan + 2, out_chan,
kernel_size, stride=stride, padding=padding, dilation=dilation,
groups=groups, bias=bias)
def forward(self, x):
batchsize, H, W = x.size(0), x.size(2), x.size(3)
h_range = torch.linspace(-1, 1, H, device=x.device, dtype=x.dtype)
w_range = torch.linspace(-1, 1, W, device=x.device, dtype=x.dtype)
h_chan, w_chan = torch.meshgrid(h_range, w_range)
h_chan = h_chan.expand([batchsize, 1, -1, -1])
w_chan = w_chan.expand([batchsize, 1, -1, -1])
feat = torch.cat([h_chan, w_chan, x], dim=1)
return F.conv2d(feat, self.weight, self.bias, self.stride, self.
padding, self.dilation, self.groups)
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4, 4]
|
EncoderLayer | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/xs/cxsdhgqd6vd2b7mfr2fdmpsoqbgad2hwpchzeouqh3m4xrstendj.py
# Topologically Sorted Source Nodes: [mul, x_att], Original ATen: [aten.mul, aten.native_layer_norm]
# Source node to ATen node mapping:
# mul => mul
# x_att => var_mean
# Graph fragment:
# %mul : [num_users=2] = call_function[target=torch.ops.aten.mul.Tensor](args = (%primals_1, %primals_2), kwargs = {})
# %var_mean : [num_users=2] = call_function[target=torch.ops.aten.var_mean.correction](args = (%mul, [2]), kwargs = {correction: 0, keepdim: True})
triton_poi_fused_mul_native_layer_norm_0 = async_compile.triton('triton_poi_fused_mul_native_layer_norm_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_mul_native_layer_norm_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 8, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_mul_native_layer_norm_0(in_ptr0, in_ptr1, out_ptr0, out_ptr1, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (4*x0), xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (4*x0), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr0 + (1 + (4*x0)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr1 + (1 + (4*x0)), xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr0 + (2 + (4*x0)), xmask, eviction_policy='evict_last')
tmp8 = tl.load(in_ptr1 + (2 + (4*x0)), xmask, eviction_policy='evict_last')
tmp11 = tl.load(in_ptr0 + (3 + (4*x0)), xmask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr1 + (3 + (4*x0)), xmask, eviction_policy='evict_last')
tmp2 = tmp0 * tmp1
tmp5 = tmp3 * tmp4
tmp6 = tmp2 + tmp5
tmp9 = tmp7 * tmp8
tmp10 = tmp6 + tmp9
tmp13 = tmp11 * tmp12
tmp14 = tmp10 + tmp13
tmp15 = 4.0
tmp16 = tmp14 / tmp15
tmp17 = tmp2 - tmp16
tmp18 = tmp17 * tmp17
tmp19 = tmp5 - tmp16
tmp20 = tmp19 * tmp19
tmp21 = tmp18 + tmp20
tmp22 = tmp9 - tmp16
tmp23 = tmp22 * tmp22
tmp24 = tmp21 + tmp23
tmp25 = tmp13 - tmp16
tmp26 = tmp25 * tmp25
tmp27 = tmp24 + tmp26
tmp28 = tmp27 / tmp15
tl.store(out_ptr0 + (x0), tmp16, xmask)
tl.store(out_ptr1 + (x0), tmp28, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/br/cbrp2mksal5nzeuhv7d2l7c7vwkkfy7luple4fdf3knpt4yyduzw.py
# Topologically Sorted Source Nodes: [mul, x_att], Original ATen: [aten.mul, aten.native_layer_norm]
# Source node to ATen node mapping:
# mul => mul
# x_att => add, add_1, mul_1, mul_2, rsqrt, sub
# Graph fragment:
# %mul : [num_users=2] = call_function[target=torch.ops.aten.mul.Tensor](args = (%primals_1, %primals_2), kwargs = {})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%getitem, 1e-05), kwargs = {})
# %rsqrt : [num_users=1] = call_function[target=torch.ops.aten.rsqrt.default](args = (%add,), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%mul, %getitem_1), kwargs = {})
# %mul_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sub, %rsqrt), kwargs = {})
# %mul_2 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%mul_1, %primals_3), kwargs = {})
# %add_1 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul_2, %primals_4), kwargs = {})
triton_poi_fused_mul_native_layer_norm_1 = async_compile.triton('triton_poi_fused_mul_native_layer_norm_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: '*fp32', 5: '*fp32', 6: '*fp32', 7: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5, 6, 7), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_mul_native_layer_norm_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 6, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_mul_native_layer_norm_1(in_ptr0, in_ptr1, in_ptr2, in_ptr3, in_ptr4, in_ptr5, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
x0 = xindex % 4
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr1 + (x2), xmask)
tmp3 = tl.load(in_ptr2 + (x1), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr3 + (x1), xmask, eviction_policy='evict_last')
tmp10 = tl.load(in_ptr4 + (x0), xmask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr5 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 * tmp1
tmp4 = tmp2 - tmp3
tmp6 = 1e-05
tmp7 = tmp5 + tmp6
tmp8 = libdevice.rsqrt(tmp7)
tmp9 = tmp4 * tmp8
tmp11 = tmp9 * tmp10
tmp13 = tmp11 + tmp12
tl.store(out_ptr0 + (x2), tmp13, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/7k/c7kkxqo5r65gqykuvge3exgf3trgxmm4raf7gypitw4ynuylbeao.py
# Topologically Sorted Source Nodes: [matmul], Original ATen: [aten.clone]
# Source node to ATen node mapping:
# matmul => clone
# Graph fragment:
# %clone : [num_users=1] = call_function[target=torch.ops.aten.clone.default](args = (%expand,), kwargs = {memory_format: torch.contiguous_format})
triton_poi_fused_clone_2 = async_compile.triton('triton_poi_fused_clone_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16, 4], tile_hint=TileHint.DEFAULT,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clone_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clone_2(in_ptr0, in_ptr1, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = (yindex // 4)
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + (4*x2) + (16*y1)), xmask & ymask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (y0), ymask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(out_ptr0 + (x2 + (4*y3)), tmp2, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/jx/cjxnq7ke3cxy6jycsbwziiw2ic65zjzt43fodlp6vavjnqmbhbvt.py
# Topologically Sorted Source Nodes: [scores, eq, scores_1, p_attn], Original ATen: [aten.div, aten.eq, aten.masked_fill, aten._softmax]
# Source node to ATen node mapping:
# eq => eq
# p_attn => amax, exp, sub_1, sum_1
# scores => div
# scores_1 => full_default, where
# Graph fragment:
# %div : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%view_11, 1.0), kwargs = {})
# %eq : [num_users=1] = call_function[target=torch.ops.aten.eq.Scalar](args = (%unsqueeze, 0), kwargs = {})
# %full_default : [num_users=1] = call_function[target=torch.ops.aten.full.default](args = ([], -1000000000.0), kwargs = {dtype: torch.float32, layout: torch.strided, device: cuda:0, pin_memory: False})
# %where : [num_users=2] = call_function[target=torch.ops.aten.where.self](args = (%eq, %full_default, %div), kwargs = {})
# %amax : [num_users=1] = call_function[target=torch.ops.aten.amax.default](args = (%where, [-1], True), kwargs = {})
# %sub_1 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%where, %amax), kwargs = {})
# %exp : [num_users=2] = call_function[target=torch.ops.aten.exp.default](args = (%sub_1,), kwargs = {})
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%exp, [-1], True), kwargs = {})
triton_poi_fused__softmax_div_eq_masked_fill_3 = async_compile.triton('triton_poi_fused__softmax_div_eq_masked_fill_3', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_div_eq_masked_fill_3', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 8, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_div_eq_masked_fill_3(in_ptr0, in_ptr1, out_ptr0, out_ptr1, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 4
x2 = (xindex // 16)
x3 = xindex
tmp0 = tl.load(in_ptr0 + ((4*x0) + (16*x2)), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr1 + (4*x3), xmask, eviction_policy='evict_last')
tmp8 = tl.load(in_ptr0 + (1 + (4*x0) + (16*x2)), xmask, eviction_policy='evict_last')
tmp10 = tl.load(in_ptr1 + (1 + (4*x3)), xmask, eviction_policy='evict_last')
tmp14 = tl.load(in_ptr0 + (2 + (4*x0) + (16*x2)), xmask, eviction_policy='evict_last')
tmp16 = tl.load(in_ptr1 + (2 + (4*x3)), xmask, eviction_policy='evict_last')
tmp20 = tl.load(in_ptr0 + (3 + (4*x0) + (16*x2)), xmask, eviction_policy='evict_last')
tmp22 = tl.load(in_ptr1 + (3 + (4*x3)), xmask, eviction_policy='evict_last')
tmp1 = 0.0
tmp2 = tmp0 == tmp1
tmp4 = 1.0
tmp5 = tmp3 * tmp4
tmp6 = -1000000000.0
tmp7 = tl.where(tmp2, tmp6, tmp5)
tmp9 = tmp8 == tmp1
tmp11 = tmp10 * tmp4
tmp12 = tl.where(tmp9, tmp6, tmp11)
tmp13 = triton_helpers.maximum(tmp7, tmp12)
tmp15 = tmp14 == tmp1
tmp17 = tmp16 * tmp4
tmp18 = tl.where(tmp15, tmp6, tmp17)
tmp19 = triton_helpers.maximum(tmp13, tmp18)
tmp21 = tmp20 == tmp1
tmp23 = tmp22 * tmp4
tmp24 = tl.where(tmp21, tmp6, tmp23)
tmp25 = triton_helpers.maximum(tmp19, tmp24)
tmp26 = tmp7 - tmp25
tmp27 = tl_math.exp(tmp26)
tmp28 = tmp12 - tmp25
tmp29 = tl_math.exp(tmp28)
tmp30 = tmp27 + tmp29
tmp31 = tmp18 - tmp25
tmp32 = tl_math.exp(tmp31)
tmp33 = tmp30 + tmp32
tmp34 = tmp24 - tmp25
tmp35 = tl_math.exp(tmp34)
tmp36 = tmp33 + tmp35
tl.store(out_ptr0 + (x3), tmp25, xmask)
tl.store(out_ptr1 + (x3), tmp36, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/vi/cvij62mui45ampgewm2umhno5rymmzocalwymmjzsoqzlhv4a2u3.py
# Topologically Sorted Source Nodes: [scores, eq, scores_1, p_attn], Original ATen: [aten.div, aten.eq, aten.masked_fill, aten._softmax]
# Source node to ATen node mapping:
# eq => eq
# p_attn => amax, div_1, exp, sub_1
# scores => div
# scores_1 => full_default, where
# Graph fragment:
# %div : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%view_11, 1.0), kwargs = {})
# %eq : [num_users=1] = call_function[target=torch.ops.aten.eq.Scalar](args = (%unsqueeze, 0), kwargs = {})
# %full_default : [num_users=1] = call_function[target=torch.ops.aten.full.default](args = ([], -1000000000.0), kwargs = {dtype: torch.float32, layout: torch.strided, device: cuda:0, pin_memory: False})
# %where : [num_users=2] = call_function[target=torch.ops.aten.where.self](args = (%eq, %full_default, %div), kwargs = {})
# %amax : [num_users=1] = call_function[target=torch.ops.aten.amax.default](args = (%where, [-1], True), kwargs = {})
# %sub_1 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%where, %amax), kwargs = {})
# %exp : [num_users=2] = call_function[target=torch.ops.aten.exp.default](args = (%sub_1,), kwargs = {})
# %div_1 : [num_users=2] = call_function[target=torch.ops.aten.div.Tensor](args = (%exp, %sum_1), kwargs = {})
triton_poi_fused__softmax_div_eq_masked_fill_4 = async_compile.triton('triton_poi_fused__softmax_div_eq_masked_fill_4', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_div_eq_masked_fill_4', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 4, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_div_eq_masked_fill_4(in_out_ptr0, in_ptr0, in_ptr1, in_ptr2, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = (xindex // 64)
x4 = xindex % 16
x5 = xindex
x6 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x4 + (16*x3)), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_out_ptr0 + (x5), xmask)
tmp8 = tl.load(in_ptr1 + (x6), xmask, eviction_policy='evict_last')
tmp11 = tl.load(in_ptr2 + (x6), xmask, eviction_policy='evict_last')
tmp1 = 0.0
tmp2 = tmp0 == tmp1
tmp4 = 1.0
tmp5 = tmp3 * tmp4
tmp6 = -1000000000.0
tmp7 = tl.where(tmp2, tmp6, tmp5)
tmp9 = tmp7 - tmp8
tmp10 = tl_math.exp(tmp9)
tmp12 = tmp10 / tmp11
tl.store(in_out_ptr0 + (x5), tmp12, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/we/cwe54p4p4jvwbdktkpj3wy2coheu6f3r3dgvi7ozm7xjfk4mgbwx.py
# Topologically Sorted Source Nodes: [contiguous], Original ATen: [aten.clone]
# Source node to ATen node mapping:
# contiguous => clone_4
# Graph fragment:
# %clone_4 : [num_users=1] = call_function[target=torch.ops.aten.clone.default](args = (%permute_7,), kwargs = {memory_format: torch.contiguous_format})
triton_poi_fused_clone_5 = async_compile.triton('triton_poi_fused_clone_5', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16, 4], tile_hint=TileHint.SQUARE,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clone_5', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clone_5(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = (yindex // 4)
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + (4*x2) + (16*y1)), xmask & ymask, eviction_policy='evict_last')
tl.store(out_ptr0 + (x2 + (4*y3)), tmp0, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/5l/c5lfswxa2a4aybcgsyhcvcycmhp5ssxucoryh4vi63bvhownmzup.py
# Topologically Sorted Source Nodes: [x_2, mul_1, x_affine], Original ATen: [aten.add, aten.mul, aten.native_layer_norm]
# Source node to ATen node mapping:
# mul_1 => mul_3
# x_2 => add_2
# x_affine => var_mean_1
# Graph fragment:
# %add_2 : [num_users=2] = call_function[target=torch.ops.aten.add.Tensor](args = (%primals_1, %view_17), kwargs = {})
# %mul_3 : [num_users=2] = call_function[target=torch.ops.aten.mul.Tensor](args = (%add_2, %primals_2), kwargs = {})
# %var_mean_1 : [num_users=2] = call_function[target=torch.ops.aten.var_mean.correction](args = (%mul_3, [2]), kwargs = {correction: 0, keepdim: True})
triton_poi_fused_add_mul_native_layer_norm_6 = async_compile.triton('triton_poi_fused_add_mul_native_layer_norm_6', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: '*fp32', 5: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_mul_native_layer_norm_6', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 12, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_mul_native_layer_norm_6(in_ptr0, in_ptr1, in_ptr2, out_ptr0, out_ptr1, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (4*x0), xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (4*x0), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr2 + (4*x0), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (1 + (4*x0)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr1 + (1 + (4*x0)), xmask, eviction_policy='evict_last')
tmp8 = tl.load(in_ptr2 + (1 + (4*x0)), xmask, eviction_policy='evict_last')
tmp11 = tl.load(in_ptr0 + (2 + (4*x0)), xmask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr1 + (2 + (4*x0)), xmask, eviction_policy='evict_last')
tmp14 = tl.load(in_ptr2 + (2 + (4*x0)), xmask, eviction_policy='evict_last')
tmp17 = tl.load(in_ptr0 + (3 + (4*x0)), xmask, eviction_policy='evict_last')
tmp18 = tl.load(in_ptr1 + (3 + (4*x0)), xmask, eviction_policy='evict_last')
tmp20 = tl.load(in_ptr2 + (3 + (4*x0)), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp4 = tmp2 * tmp3
tmp7 = tmp5 + tmp6
tmp9 = tmp7 * tmp8
tmp10 = tmp4 + tmp9
tmp13 = tmp11 + tmp12
tmp15 = tmp13 * tmp14
tmp16 = tmp10 + tmp15
tmp19 = tmp17 + tmp18
tmp21 = tmp19 * tmp20
tmp22 = tmp16 + tmp21
tmp23 = 4.0
tmp24 = tmp22 / tmp23
tmp25 = tmp4 - tmp24
tmp26 = tmp25 * tmp25
tmp27 = tmp9 - tmp24
tmp28 = tmp27 * tmp27
tmp29 = tmp26 + tmp28
tmp30 = tmp15 - tmp24
tmp31 = tmp30 * tmp30
tmp32 = tmp29 + tmp31
tmp33 = tmp21 - tmp24
tmp34 = tmp33 * tmp33
tmp35 = tmp32 + tmp34
tmp36 = tmp35 / tmp23
tl.store(out_ptr0 + (x0), tmp24, xmask)
tl.store(out_ptr1 + (x0), tmp36, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/az/cazk6ygry7rdh7hlohmxpdos4ijkax24xrpnhplpr2hmjlwa64s4.py
# Topologically Sorted Source Nodes: [x_2, mul_1, x_affine], Original ATen: [aten.add, aten.mul, aten.native_layer_norm]
# Source node to ATen node mapping:
# mul_1 => mul_3
# x_2 => add_2
# x_affine => add_3, add_4, mul_4, mul_5, rsqrt_1, sub_2
# Graph fragment:
# %add_2 : [num_users=2] = call_function[target=torch.ops.aten.add.Tensor](args = (%primals_1, %view_17), kwargs = {})
# %mul_3 : [num_users=2] = call_function[target=torch.ops.aten.mul.Tensor](args = (%add_2, %primals_2), kwargs = {})
# %add_3 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%getitem_2, 1e-05), kwargs = {})
# %rsqrt_1 : [num_users=1] = call_function[target=torch.ops.aten.rsqrt.default](args = (%add_3,), kwargs = {})
# %sub_2 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%mul_3, %getitem_3), kwargs = {})
# %mul_4 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sub_2, %rsqrt_1), kwargs = {})
# %mul_5 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%mul_4, %primals_13), kwargs = {})
# %add_4 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul_5, %primals_14), kwargs = {})
triton_poi_fused_add_mul_native_layer_norm_7 = async_compile.triton('triton_poi_fused_add_mul_native_layer_norm_7', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: '*fp32', 5: '*fp32', 6: '*fp32', 7: '*fp32', 8: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5, 6, 7, 8), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_mul_native_layer_norm_7', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 7, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_mul_native_layer_norm_7(in_ptr0, in_ptr1, in_ptr2, in_ptr3, in_ptr4, in_ptr5, in_ptr6, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
x0 = xindex % 4
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr1 + (x2), xmask)
tmp3 = tl.load(in_ptr2 + (x2), xmask)
tmp5 = tl.load(in_ptr3 + (x1), xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr4 + (x1), xmask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr5 + (x0), xmask, eviction_policy='evict_last')
tmp14 = tl.load(in_ptr6 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp4 = tmp2 * tmp3
tmp6 = tmp4 - tmp5
tmp8 = 1e-05
tmp9 = tmp7 + tmp8
tmp10 = libdevice.rsqrt(tmp9)
tmp11 = tmp6 * tmp10
tmp13 = tmp11 * tmp12
tmp15 = tmp13 + tmp14
tl.store(out_ptr0 + (x2), tmp15, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/23/c23j2vk753qpctrm5kblwdo7f2zh4pnjylmlrcdhyl7syfjonudr.py
# Topologically Sorted Source Nodes: [relu], Original ATen: [aten.relu, aten.threshold_backward]
# Source node to ATen node mapping:
# relu => relu
# Graph fragment:
# %relu : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%view_19,), kwargs = {})
# %le : [num_users=1] = call_function[target=torch.ops.aten.le.Scalar](args = (%relu, 0), kwargs = {})
triton_poi_fused_relu_threshold_backward_8 = async_compile.triton('triton_poi_fused_relu_threshold_backward_8', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*i1', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_relu_threshold_backward_8', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_relu_threshold_backward_8(in_out_ptr0, in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(in_out_ptr0 + (x2), tmp4, xmask)
tl.store(out_ptr0 + (x2), tmp6, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/x7/cx7y3kua2bihmqz24vhzuizsgkwgtdcyxynklnn7saro6oxogdky.py
# Topologically Sorted Source Nodes: [x_2, add_1], Original ATen: [aten.add]
# Source node to ATen node mapping:
# add_1 => add_5
# x_2 => add_2
# Graph fragment:
# %add_2 : [num_users=2] = call_function[target=torch.ops.aten.add.Tensor](args = (%primals_1, %view_17), kwargs = {})
# %add_5 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%add_2, %view_21), kwargs = {})
triton_poi_fused_add_9 = async_compile.triton('triton_poi_fused_add_9', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_9', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 4, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_9(in_out_ptr0, in_ptr0, in_ptr1, in_ptr2, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr1 + (x2), xmask)
tmp3 = tl.load(in_out_ptr0 + (x2), xmask)
tmp4 = tl.load(in_ptr2 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp5 = tmp3 + tmp4
tmp6 = tmp2 + tmp5
tl.store(in_out_ptr0 + (x2), tmp6, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9, primals_10, primals_11, primals_12, primals_13, primals_14, primals_15, primals_16, primals_17, primals_18 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_2, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_3, (4, ), (1, ))
assert_size_stride(primals_4, (4, ), (1, ))
assert_size_stride(primals_5, (4, 4), (4, 1))
assert_size_stride(primals_6, (4, ), (1, ))
assert_size_stride(primals_7, (4, 4), (4, 1))
assert_size_stride(primals_8, (4, ), (1, ))
assert_size_stride(primals_9, (4, 4), (4, 1))
assert_size_stride(primals_10, (4, ), (1, ))
assert_size_stride(primals_11, (4, 4), (4, 1))
assert_size_stride(primals_12, (4, ), (1, ))
assert_size_stride(primals_13, (4, ), (1, ))
assert_size_stride(primals_14, (4, ), (1, ))
assert_size_stride(primals_15, (4, 4), (4, 1))
assert_size_stride(primals_16, (4, ), (1, ))
assert_size_stride(primals_17, (4, 4), (4, 1))
assert_size_stride(primals_18, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 1), (4, 1, 16), torch.float32)
buf1 = empty_strided_cuda((4, 4, 1), (4, 1, 16), torch.float32)
# Topologically Sorted Source Nodes: [mul, x_att], Original ATen: [aten.mul, aten.native_layer_norm]
stream0 = get_raw_stream(0)
triton_poi_fused_mul_native_layer_norm_0.run(primals_1, primals_2, buf0, buf1, 16, grid=grid(16), stream=stream0)
buf2 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [mul, x_att], Original ATen: [aten.mul, aten.native_layer_norm]
triton_poi_fused_mul_native_layer_norm_1.run(primals_1, primals_2, buf0, buf1, primals_3, primals_4, buf2, 64, grid=grid(64), stream=stream0)
del primals_3
del primals_4
buf3 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(buf2, (16, 4), (4, 1), 0), reinterpret_tensor(primals_5, (4, 4), (1, 4), 0), out=buf3)
buf4 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(buf2, (16, 4), (4, 1), 0), reinterpret_tensor(primals_7, (4, 4), (1, 4), 0), out=buf4)
buf5 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(buf2, (16, 4), (4, 1), 0), reinterpret_tensor(primals_9, (4, 4), (1, 4), 0), out=buf5)
buf6 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 1), torch.float32)
# Topologically Sorted Source Nodes: [matmul], Original ATen: [aten.clone]
triton_poi_fused_clone_2.run(buf3, primals_6, buf6, 16, 4, grid=grid(16, 4), stream=stream0)
del primals_6
buf7 = reinterpret_tensor(buf3, (4, 4, 1, 4), (16, 4, 4, 1), 0); del buf3 # reuse
# Topologically Sorted Source Nodes: [matmul], Original ATen: [aten.clone]
triton_poi_fused_clone_2.run(buf4, primals_8, buf7, 16, 4, grid=grid(16, 4), stream=stream0)
del primals_8
buf8 = empty_strided_cuda((16, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [matmul], Original ATen: [aten.bmm]
extern_kernels.bmm(reinterpret_tensor(buf6, (16, 4, 1), (4, 1, 0), 0), reinterpret_tensor(buf7, (16, 1, 4), (4, 0, 1), 0), out=buf8)
buf9 = reinterpret_tensor(buf4, (4, 4, 4, 1), (16, 4, 1, 64), 0); del buf4 # reuse
buf10 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 64), torch.float32)
# Topologically Sorted Source Nodes: [scores, eq, scores_1, p_attn], Original ATen: [aten.div, aten.eq, aten.masked_fill, aten._softmax]
triton_poi_fused__softmax_div_eq_masked_fill_3.run(primals_2, buf8, buf9, buf10, 64, grid=grid(64), stream=stream0)
buf11 = reinterpret_tensor(buf8, (4, 4, 4, 4), (64, 16, 4, 1), 0); del buf8 # reuse
# Topologically Sorted Source Nodes: [scores, eq, scores_1, p_attn], Original ATen: [aten.div, aten.eq, aten.masked_fill, aten._softmax]
triton_poi_fused__softmax_div_eq_masked_fill_4.run(buf11, primals_2, buf9, buf10, 256, grid=grid(256), stream=stream0)
buf12 = reinterpret_tensor(buf9, (4, 4, 4, 1), (16, 4, 1, 1), 0); del buf9 # reuse
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.clone]
triton_poi_fused_clone_2.run(buf5, primals_10, buf12, 16, 4, grid=grid(16, 4), stream=stream0)
del primals_10
buf13 = reinterpret_tensor(buf5, (16, 4, 1), (4, 1, 1), 0); del buf5 # reuse
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.bmm]
extern_kernels.bmm(reinterpret_tensor(buf11, (16, 4, 4), (16, 4, 1), 0), reinterpret_tensor(buf12, (16, 4, 1), (4, 1, 0), 0), out=buf13)
buf14 = reinterpret_tensor(buf10, (4, 4, 4, 1), (16, 4, 1, 1), 0); del buf10 # reuse
# Topologically Sorted Source Nodes: [contiguous], Original ATen: [aten.clone]
triton_poi_fused_clone_5.run(buf13, buf14, 16, 4, grid=grid(16, 4), stream=stream0)
buf15 = reinterpret_tensor(buf13, (16, 4), (4, 1), 0); del buf13 # reuse
# Topologically Sorted Source Nodes: [x_att_1], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_12, reinterpret_tensor(buf14, (16, 4), (4, 1), 0), reinterpret_tensor(primals_11, (4, 4), (1, 4), 0), alpha=1, beta=1, out=buf15)
del primals_12
buf16 = buf1; del buf1 # reuse
buf17 = buf0; del buf0 # reuse
# Topologically Sorted Source Nodes: [x_2, mul_1, x_affine], Original ATen: [aten.add, aten.mul, aten.native_layer_norm]
triton_poi_fused_add_mul_native_layer_norm_6.run(primals_1, buf15, primals_2, buf16, buf17, 16, grid=grid(16), stream=stream0)
buf18 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [x_2, mul_1, x_affine], Original ATen: [aten.add, aten.mul, aten.native_layer_norm]
triton_poi_fused_add_mul_native_layer_norm_7.run(primals_1, buf15, primals_2, buf16, buf17, primals_13, primals_14, buf18, 64, grid=grid(64), stream=stream0)
del buf16
del buf17
del primals_14
buf19 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(buf18, (16, 4), (4, 1), 0), reinterpret_tensor(primals_15, (4, 4), (1, 4), 0), out=buf19)
buf20 = reinterpret_tensor(buf19, (4, 4, 4), (16, 4, 1), 0); del buf19 # reuse
buf23 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.bool)
# Topologically Sorted Source Nodes: [relu], Original ATen: [aten.relu, aten.threshold_backward]
triton_poi_fused_relu_threshold_backward_8.run(buf20, primals_16, buf23, 64, grid=grid(64), stream=stream0)
del primals_16
buf21 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(buf20, (16, 4), (4, 1), 0), reinterpret_tensor(primals_17, (4, 4), (1, 4), 0), out=buf21)
buf22 = reinterpret_tensor(buf21, (4, 4, 4), (16, 4, 1), 0); del buf21 # reuse
# Topologically Sorted Source Nodes: [x_2, add_1], Original ATen: [aten.add]
triton_poi_fused_add_9.run(buf22, primals_1, buf15, primals_18, 64, grid=grid(64), stream=stream0)
del primals_18
return (buf22, primals_1, primals_2, primals_13, reinterpret_tensor(buf2, (16, 4), (4, 1), 0), buf11, reinterpret_tensor(buf14, (16, 4), (4, 1), 0), buf15, reinterpret_tensor(buf18, (16, 4), (4, 1), 0), reinterpret_tensor(buf20, (16, 4), (4, 1), 0), primals_17, buf23, primals_15, primals_11, reinterpret_tensor(buf12, (16, 1, 4), (4, 1, 1), 0), reinterpret_tensor(buf6, (16, 1, 4), (4, 1, 1), 0), reinterpret_tensor(buf7, (16, 4, 1), (4, 1, 4), 0), primals_9, primals_7, primals_5, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_8 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_9 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_10 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_11 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_12 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_13 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_14 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_15 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_16 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_17 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_18 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9, primals_10, primals_11, primals_12, primals_13, primals_14, primals_15, primals_16, primals_17, primals_18])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import math
import torch
import torch.nn as nn
import torch.nn.functional as F
class AffineLayer(nn.Module):
def __init__(self, dropout, d_model, d_ff):
super(AffineLayer, self).__init__()
self.w_1 = nn.Linear(d_model, d_ff)
self.w_2 = nn.Linear(d_ff, d_model)
self.dropout = nn.Dropout(dropout)
def forward(self, x):
return self.w_2(self.dropout(F.relu(self.w_1(x))))
class MultiHeadedAttention(nn.Module):
def __init__(self, num_head, d_model, dropout=0.1):
super(MultiHeadedAttention, self).__init__()
assert d_model % num_head == 0
self.d_k = d_model // num_head
self.h = num_head
self.linear_key = nn.Linear(d_model, d_model)
self.linear_value = nn.Linear(d_model, d_model)
self.linear_query = nn.Linear(d_model, d_model)
self.linear_out = nn.Linear(d_model, d_model)
self.dropout = nn.Dropout(p=dropout)
def attention(self, query, key, value, mask, dropout=None):
d_k = query.size(-1)
scores = torch.matmul(query, key.transpose(-2, -1)) / math.sqrt(d_k)
scores = scores.masked_fill(mask == 0, -1000000000.0)
p_attn = F.softmax(scores, dim=-1)
if dropout is not None:
p_attn = dropout(p_attn)
return torch.matmul(p_attn, value), p_attn
def forward(self, query, key, value, mask):
nbatches = query.size(0)
query = self.linear_query(query).view(nbatches, -1, self.h, self.d_k
).transpose(1, 2)
key = self.linear_key(key).view(nbatches, -1, self.h, self.d_k
).transpose(1, 2)
value = self.linear_value(value).view(nbatches, -1, self.h, self.d_k
).transpose(1, 2)
mask = mask.unsqueeze(1)
x, _attn = self.attention(query, key, value, mask, dropout=self.dropout
)
x = x.transpose(1, 2).contiguous().view(nbatches, -1, self.h * self.d_k
)
return self.linear_out(x)
class EncoderLayer(nn.Module):
def __init__(self, num_head, dropout, d_model, d_ff):
super(EncoderLayer, self).__init__()
self.att_layer = MultiHeadedAttention(num_head, d_model, dropout)
self.norm_att = nn.LayerNorm(d_model)
self.dropout_att = nn.Dropout(dropout)
self.affine_layer = AffineLayer(dropout, d_model, d_ff)
self.norm_affine = nn.LayerNorm(d_model)
self.dropout_affine = nn.Dropout(dropout)
def forward(self, x, mask):
x_att = self.norm_att(x * mask)
x_att = self.att_layer(x_att, x_att, x_att, mask)
x = x + self.dropout_att(x_att)
x_affine = self.norm_affine(x * mask)
x_affine = self.affine_layer(x_affine)
return x + self.dropout_affine(x_affine)
def get_inputs():
return [torch.rand([4, 4, 4]), torch.rand([4, 4, 4])]
def get_init_inputs():
return [[], {'num_head': 4, 'dropout': 0.5, 'd_model': 4, 'd_ff': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
import math
import torch.nn as nn
import torch.nn.functional as F
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_mul_native_layer_norm_0(in_ptr0, in_ptr1, out_ptr0,
out_ptr1, xnumel, XBLOCK: tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + 4 * x0, xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + 4 * x0, xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr0 + (1 + 4 * x0), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr1 + (1 + 4 * x0), xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr0 + (2 + 4 * x0), xmask, eviction_policy='evict_last')
tmp8 = tl.load(in_ptr1 + (2 + 4 * x0), xmask, eviction_policy='evict_last')
tmp11 = tl.load(in_ptr0 + (3 + 4 * x0), xmask, eviction_policy='evict_last'
)
tmp12 = tl.load(in_ptr1 + (3 + 4 * x0), xmask, eviction_policy='evict_last'
)
tmp2 = tmp0 * tmp1
tmp5 = tmp3 * tmp4
tmp6 = tmp2 + tmp5
tmp9 = tmp7 * tmp8
tmp10 = tmp6 + tmp9
tmp13 = tmp11 * tmp12
tmp14 = tmp10 + tmp13
tmp15 = 4.0
tmp16 = tmp14 / tmp15
tmp17 = tmp2 - tmp16
tmp18 = tmp17 * tmp17
tmp19 = tmp5 - tmp16
tmp20 = tmp19 * tmp19
tmp21 = tmp18 + tmp20
tmp22 = tmp9 - tmp16
tmp23 = tmp22 * tmp22
tmp24 = tmp21 + tmp23
tmp25 = tmp13 - tmp16
tmp26 = tmp25 * tmp25
tmp27 = tmp24 + tmp26
tmp28 = tmp27 / tmp15
tl.store(out_ptr0 + x0, tmp16, xmask)
tl.store(out_ptr1 + x0, tmp28, xmask)
@triton.jit
def triton_poi_fused_mul_native_layer_norm_1(in_ptr0, in_ptr1, in_ptr2,
in_ptr3, in_ptr4, in_ptr5, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
x0 = xindex % 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr1 + x2, xmask)
tmp3 = tl.load(in_ptr2 + x1, xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr3 + x1, xmask, eviction_policy='evict_last')
tmp10 = tl.load(in_ptr4 + x0, xmask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr5 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 * tmp1
tmp4 = tmp2 - tmp3
tmp6 = 1e-05
tmp7 = tmp5 + tmp6
tmp8 = libdevice.rsqrt(tmp7)
tmp9 = tmp4 * tmp8
tmp11 = tmp9 * tmp10
tmp13 = tmp11 + tmp12
tl.store(out_ptr0 + x2, tmp13, xmask)
@triton.jit
def triton_poi_fused_clone_2(in_ptr0, in_ptr1, out_ptr0, ynumel, xnumel,
YBLOCK: tl.constexpr, XBLOCK: tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = yindex // 4
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + 4 * x2 + 16 * y1), xmask & ymask,
eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + y0, ymask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(out_ptr0 + (x2 + 4 * y3), tmp2, xmask & ymask)
@triton.jit
def triton_poi_fused__softmax_div_eq_masked_fill_3(in_ptr0, in_ptr1,
out_ptr0, out_ptr1, xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 4
x2 = xindex // 16
x3 = xindex
tmp0 = tl.load(in_ptr0 + (4 * x0 + 16 * x2), xmask, eviction_policy=
'evict_last')
tmp3 = tl.load(in_ptr1 + 4 * x3, xmask, eviction_policy='evict_last')
tmp8 = tl.load(in_ptr0 + (1 + 4 * x0 + 16 * x2), xmask, eviction_policy
='evict_last')
tmp10 = tl.load(in_ptr1 + (1 + 4 * x3), xmask, eviction_policy='evict_last'
)
tmp14 = tl.load(in_ptr0 + (2 + 4 * x0 + 16 * x2), xmask,
eviction_policy='evict_last')
tmp16 = tl.load(in_ptr1 + (2 + 4 * x3), xmask, eviction_policy='evict_last'
)
tmp20 = tl.load(in_ptr0 + (3 + 4 * x0 + 16 * x2), xmask,
eviction_policy='evict_last')
tmp22 = tl.load(in_ptr1 + (3 + 4 * x3), xmask, eviction_policy='evict_last'
)
tmp1 = 0.0
tmp2 = tmp0 == tmp1
tmp4 = 1.0
tmp5 = tmp3 * tmp4
tmp6 = -1000000000.0
tmp7 = tl.where(tmp2, tmp6, tmp5)
tmp9 = tmp8 == tmp1
tmp11 = tmp10 * tmp4
tmp12 = tl.where(tmp9, tmp6, tmp11)
tmp13 = triton_helpers.maximum(tmp7, tmp12)
tmp15 = tmp14 == tmp1
tmp17 = tmp16 * tmp4
tmp18 = tl.where(tmp15, tmp6, tmp17)
tmp19 = triton_helpers.maximum(tmp13, tmp18)
tmp21 = tmp20 == tmp1
tmp23 = tmp22 * tmp4
tmp24 = tl.where(tmp21, tmp6, tmp23)
tmp25 = triton_helpers.maximum(tmp19, tmp24)
tmp26 = tmp7 - tmp25
tmp27 = tl_math.exp(tmp26)
tmp28 = tmp12 - tmp25
tmp29 = tl_math.exp(tmp28)
tmp30 = tmp27 + tmp29
tmp31 = tmp18 - tmp25
tmp32 = tl_math.exp(tmp31)
tmp33 = tmp30 + tmp32
tmp34 = tmp24 - tmp25
tmp35 = tl_math.exp(tmp34)
tmp36 = tmp33 + tmp35
tl.store(out_ptr0 + x3, tmp25, xmask)
tl.store(out_ptr1 + x3, tmp36, xmask)
@triton.jit
def triton_poi_fused__softmax_div_eq_masked_fill_4(in_out_ptr0, in_ptr0,
in_ptr1, in_ptr2, xnumel, XBLOCK: tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex // 64
x4 = xindex % 16
x5 = xindex
x6 = xindex // 4
tmp0 = tl.load(in_ptr0 + (x4 + 16 * x3), xmask, eviction_policy=
'evict_last')
tmp3 = tl.load(in_out_ptr0 + x5, xmask)
tmp8 = tl.load(in_ptr1 + x6, xmask, eviction_policy='evict_last')
tmp11 = tl.load(in_ptr2 + x6, xmask, eviction_policy='evict_last')
tmp1 = 0.0
tmp2 = tmp0 == tmp1
tmp4 = 1.0
tmp5 = tmp3 * tmp4
tmp6 = -1000000000.0
tmp7 = tl.where(tmp2, tmp6, tmp5)
tmp9 = tmp7 - tmp8
tmp10 = tl_math.exp(tmp9)
tmp12 = tmp10 / tmp11
tl.store(in_out_ptr0 + x5, tmp12, xmask)
@triton.jit
def triton_poi_fused_clone_5(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK: tl.
constexpr, XBLOCK: tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = yindex // 4
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + 4 * x2 + 16 * y1), xmask & ymask,
eviction_policy='evict_last')
tl.store(out_ptr0 + (x2 + 4 * y3), tmp0, xmask & ymask)
@triton.jit
def triton_poi_fused_add_mul_native_layer_norm_6(in_ptr0, in_ptr1, in_ptr2,
out_ptr0, out_ptr1, xnumel, XBLOCK: tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + 4 * x0, xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + 4 * x0, xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr2 + 4 * x0, xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (1 + 4 * x0), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr1 + (1 + 4 * x0), xmask, eviction_policy='evict_last')
tmp8 = tl.load(in_ptr2 + (1 + 4 * x0), xmask, eviction_policy='evict_last')
tmp11 = tl.load(in_ptr0 + (2 + 4 * x0), xmask, eviction_policy='evict_last'
)
tmp12 = tl.load(in_ptr1 + (2 + 4 * x0), xmask, eviction_policy='evict_last'
)
tmp14 = tl.load(in_ptr2 + (2 + 4 * x0), xmask, eviction_policy='evict_last'
)
tmp17 = tl.load(in_ptr0 + (3 + 4 * x0), xmask, eviction_policy='evict_last'
)
tmp18 = tl.load(in_ptr1 + (3 + 4 * x0), xmask, eviction_policy='evict_last'
)
tmp20 = tl.load(in_ptr2 + (3 + 4 * x0), xmask, eviction_policy='evict_last'
)
tmp2 = tmp0 + tmp1
tmp4 = tmp2 * tmp3
tmp7 = tmp5 + tmp6
tmp9 = tmp7 * tmp8
tmp10 = tmp4 + tmp9
tmp13 = tmp11 + tmp12
tmp15 = tmp13 * tmp14
tmp16 = tmp10 + tmp15
tmp19 = tmp17 + tmp18
tmp21 = tmp19 * tmp20
tmp22 = tmp16 + tmp21
tmp23 = 4.0
tmp24 = tmp22 / tmp23
tmp25 = tmp4 - tmp24
tmp26 = tmp25 * tmp25
tmp27 = tmp9 - tmp24
tmp28 = tmp27 * tmp27
tmp29 = tmp26 + tmp28
tmp30 = tmp15 - tmp24
tmp31 = tmp30 * tmp30
tmp32 = tmp29 + tmp31
tmp33 = tmp21 - tmp24
tmp34 = tmp33 * tmp33
tmp35 = tmp32 + tmp34
tmp36 = tmp35 / tmp23
tl.store(out_ptr0 + x0, tmp24, xmask)
tl.store(out_ptr1 + x0, tmp36, xmask)
@triton.jit
def triton_poi_fused_add_mul_native_layer_norm_7(in_ptr0, in_ptr1, in_ptr2,
in_ptr3, in_ptr4, in_ptr5, in_ptr6, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
x0 = xindex % 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr1 + x2, xmask)
tmp3 = tl.load(in_ptr2 + x2, xmask)
tmp5 = tl.load(in_ptr3 + x1, xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr4 + x1, xmask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr5 + x0, xmask, eviction_policy='evict_last')
tmp14 = tl.load(in_ptr6 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp4 = tmp2 * tmp3
tmp6 = tmp4 - tmp5
tmp8 = 1e-05
tmp9 = tmp7 + tmp8
tmp10 = libdevice.rsqrt(tmp9)
tmp11 = tmp6 * tmp10
tmp13 = tmp11 * tmp12
tmp15 = tmp13 + tmp14
tl.store(out_ptr0 + x2, tmp15, xmask)
@triton.jit
def triton_poi_fused_relu_threshold_backward_8(in_out_ptr0, in_ptr0,
out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(in_out_ptr0 + x2, tmp4, xmask)
tl.store(out_ptr0 + x2, tmp6, xmask)
@triton.jit
def triton_poi_fused_add_9(in_out_ptr0, in_ptr0, in_ptr1, in_ptr2, xnumel,
XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr1 + x2, xmask)
tmp3 = tl.load(in_out_ptr0 + x2, xmask)
tmp4 = tl.load(in_ptr2 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp5 = tmp3 + tmp4
tmp6 = tmp2 + tmp5
tl.store(in_out_ptr0 + x2, tmp6, xmask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7, primals_8, primals_9, primals_10, primals_11, primals_12,
primals_13, primals_14, primals_15, primals_16, primals_17, primals_18
) = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_2, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_3, (4,), (1,))
assert_size_stride(primals_4, (4,), (1,))
assert_size_stride(primals_5, (4, 4), (4, 1))
assert_size_stride(primals_6, (4,), (1,))
assert_size_stride(primals_7, (4, 4), (4, 1))
assert_size_stride(primals_8, (4,), (1,))
assert_size_stride(primals_9, (4, 4), (4, 1))
assert_size_stride(primals_10, (4,), (1,))
assert_size_stride(primals_11, (4, 4), (4, 1))
assert_size_stride(primals_12, (4,), (1,))
assert_size_stride(primals_13, (4,), (1,))
assert_size_stride(primals_14, (4,), (1,))
assert_size_stride(primals_15, (4, 4), (4, 1))
assert_size_stride(primals_16, (4,), (1,))
assert_size_stride(primals_17, (4, 4), (4, 1))
assert_size_stride(primals_18, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 1), (4, 1, 16), torch.float32)
buf1 = empty_strided_cuda((4, 4, 1), (4, 1, 16), torch.float32)
get_raw_stream(0)
triton_poi_fused_mul_native_layer_norm_0[grid(16)](primals_1,
primals_2, buf0, buf1, 16, XBLOCK=16, num_warps=1, num_stages=1)
buf2 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
triton_poi_fused_mul_native_layer_norm_1[grid(64)](primals_1,
primals_2, buf0, buf1, primals_3, primals_4, buf2, 64, XBLOCK=
64, num_warps=1, num_stages=1)
del primals_3
del primals_4
buf3 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(buf2, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_5, (4, 4), (1, 4), 0), out=buf3)
buf4 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(buf2, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_7, (4, 4), (1, 4), 0), out=buf4)
buf5 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(buf2, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_9, (4, 4), (1, 4), 0), out=buf5)
buf6 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 1), torch.float32)
triton_poi_fused_clone_2[grid(16, 4)](buf3, primals_6, buf6, 16, 4,
XBLOCK=4, YBLOCK=8, num_warps=1, num_stages=1)
del primals_6
buf7 = reinterpret_tensor(buf3, (4, 4, 1, 4), (16, 4, 4, 1), 0)
del buf3
triton_poi_fused_clone_2[grid(16, 4)](buf4, primals_8, buf7, 16, 4,
XBLOCK=4, YBLOCK=8, num_warps=1, num_stages=1)
del primals_8
buf8 = empty_strided_cuda((16, 4, 4), (16, 4, 1), torch.float32)
extern_kernels.bmm(reinterpret_tensor(buf6, (16, 4, 1), (4, 1, 0),
0), reinterpret_tensor(buf7, (16, 1, 4), (4, 0, 1), 0), out=buf8)
buf9 = reinterpret_tensor(buf4, (4, 4, 4, 1), (16, 4, 1, 64), 0)
del buf4
buf10 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 64), torch.float32)
triton_poi_fused__softmax_div_eq_masked_fill_3[grid(64)](primals_2,
buf8, buf9, buf10, 64, XBLOCK=64, num_warps=1, num_stages=1)
buf11 = reinterpret_tensor(buf8, (4, 4, 4, 4), (64, 16, 4, 1), 0)
del buf8
triton_poi_fused__softmax_div_eq_masked_fill_4[grid(256)](buf11,
primals_2, buf9, buf10, 256, XBLOCK=128, num_warps=4, num_stages=1)
buf12 = reinterpret_tensor(buf9, (4, 4, 4, 1), (16, 4, 1, 1), 0)
del buf9
triton_poi_fused_clone_2[grid(16, 4)](buf5, primals_10, buf12, 16,
4, XBLOCK=4, YBLOCK=8, num_warps=1, num_stages=1)
del primals_10
buf13 = reinterpret_tensor(buf5, (16, 4, 1), (4, 1, 1), 0)
del buf5
extern_kernels.bmm(reinterpret_tensor(buf11, (16, 4, 4), (16, 4, 1),
0), reinterpret_tensor(buf12, (16, 4, 1), (4, 1, 0), 0), out=buf13)
buf14 = reinterpret_tensor(buf10, (4, 4, 4, 1), (16, 4, 1, 1), 0)
del buf10
triton_poi_fused_clone_5[grid(16, 4)](buf13, buf14, 16, 4, XBLOCK=4,
YBLOCK=16, num_warps=1, num_stages=1)
buf15 = reinterpret_tensor(buf13, (16, 4), (4, 1), 0)
del buf13
extern_kernels.addmm(primals_12, reinterpret_tensor(buf14, (16, 4),
(4, 1), 0), reinterpret_tensor(primals_11, (4, 4), (1, 4), 0),
alpha=1, beta=1, out=buf15)
del primals_12
buf16 = buf1
del buf1
buf17 = buf0
del buf0
triton_poi_fused_add_mul_native_layer_norm_6[grid(16)](primals_1,
buf15, primals_2, buf16, buf17, 16, XBLOCK=16, num_warps=1,
num_stages=1)
buf18 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
triton_poi_fused_add_mul_native_layer_norm_7[grid(64)](primals_1,
buf15, primals_2, buf16, buf17, primals_13, primals_14, buf18,
64, XBLOCK=64, num_warps=1, num_stages=1)
del buf16
del buf17
del primals_14
buf19 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(buf18, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_15, (4, 4), (1, 4), 0), out=buf19)
buf20 = reinterpret_tensor(buf19, (4, 4, 4), (16, 4, 1), 0)
del buf19
buf23 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.bool)
triton_poi_fused_relu_threshold_backward_8[grid(64)](buf20,
primals_16, buf23, 64, XBLOCK=64, num_warps=1, num_stages=1)
del primals_16
buf21 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(buf20, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_17, (4, 4), (1, 4), 0), out=buf21)
buf22 = reinterpret_tensor(buf21, (4, 4, 4), (16, 4, 1), 0)
del buf21
triton_poi_fused_add_9[grid(64)](buf22, primals_1, buf15,
primals_18, 64, XBLOCK=64, num_warps=1, num_stages=1)
del primals_18
return buf22, primals_1, primals_2, primals_13, reinterpret_tensor(buf2,
(16, 4), (4, 1), 0), buf11, reinterpret_tensor(buf14, (16, 4), (4,
1), 0), buf15, reinterpret_tensor(buf18, (16, 4), (4, 1), 0
), reinterpret_tensor(buf20, (16, 4), (4, 1), 0
), primals_17, buf23, primals_15, primals_11, reinterpret_tensor(buf12,
(16, 1, 4), (4, 1, 1), 0), reinterpret_tensor(buf6, (16, 1, 4), (4,
1, 1), 0), reinterpret_tensor(buf7, (16, 4, 1), (4, 1, 4), 0
), primals_9, primals_7, primals_5
class AffineLayer(nn.Module):
def __init__(self, dropout, d_model, d_ff):
super(AffineLayer, self).__init__()
self.w_1 = nn.Linear(d_model, d_ff)
self.w_2 = nn.Linear(d_ff, d_model)
self.dropout = nn.Dropout(dropout)
def forward(self, x):
return self.w_2(self.dropout(F.relu(self.w_1(x))))
class MultiHeadedAttention(nn.Module):
def __init__(self, num_head, d_model, dropout=0.1):
super(MultiHeadedAttention, self).__init__()
assert d_model % num_head == 0
self.d_k = d_model // num_head
self.h = num_head
self.linear_key = nn.Linear(d_model, d_model)
self.linear_value = nn.Linear(d_model, d_model)
self.linear_query = nn.Linear(d_model, d_model)
self.linear_out = nn.Linear(d_model, d_model)
self.dropout = nn.Dropout(p=dropout)
def attention(self, query, key, value, mask, dropout=None):
d_k = query.size(-1)
scores = torch.matmul(query, key.transpose(-2, -1)) / math.sqrt(d_k)
scores = scores.masked_fill(mask == 0, -1000000000.0)
p_attn = F.softmax(scores, dim=-1)
if dropout is not None:
p_attn = dropout(p_attn)
return torch.matmul(p_attn, value), p_attn
def forward(self, query, key, value, mask):
nbatches = query.size(0)
query = self.linear_query(query).view(nbatches, -1, self.h, self.d_k
).transpose(1, 2)
key = self.linear_key(key).view(nbatches, -1, self.h, self.d_k
).transpose(1, 2)
value = self.linear_value(value).view(nbatches, -1, self.h, self.d_k
).transpose(1, 2)
mask = mask.unsqueeze(1)
x, _attn = self.attention(query, key, value, mask, dropout=self.dropout
)
x = x.transpose(1, 2).contiguous().view(nbatches, -1, self.h * self.d_k
)
return self.linear_out(x)
class EncoderLayerNew(nn.Module):
def __init__(self, num_head, dropout, d_model, d_ff):
super(EncoderLayerNew, self).__init__()
self.att_layer = MultiHeadedAttention(num_head, d_model, dropout)
self.norm_att = nn.LayerNorm(d_model)
self.dropout_att = nn.Dropout(dropout)
self.affine_layer = AffineLayer(dropout, d_model, d_ff)
self.norm_affine = nn.LayerNorm(d_model)
self.dropout_affine = nn.Dropout(dropout)
def forward(self, input_0, input_1):
primals_5 = self.att_layer.linear_key.weight
primals_3 = self.att_layer.linear_key.bias
primals_7 = self.att_layer.linear_value.weight
primals_4 = self.att_layer.linear_value.bias
primals_9 = self.att_layer.linear_query.weight
primals_6 = self.att_layer.linear_query.bias
primals_11 = self.att_layer.linear_out.weight
primals_8 = self.att_layer.linear_out.bias
primals_10 = self.norm_att.weight
primals_12 = self.norm_att.bias
primals_15 = self.affine_layer.w_1.weight
primals_13 = self.affine_layer.w_1.bias
primals_17 = self.affine_layer.w_2.weight
primals_14 = self.affine_layer.w_2.bias
primals_16 = self.norm_affine.weight
primals_18 = self.norm_affine.bias
primals_1 = input_0
primals_2 = input_1
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7, primals_8, primals_9,
primals_10, primals_11, primals_12, primals_13, primals_14,
primals_15, primals_16, primals_17, primals_18])
return output[0]
| iamxpy/pointer_summarizer | EncoderLayer | false | 6,878 | [
"Apache-2.0"
] | 1 | ebeb2ad32a45162c0da14dac0b6241b0b0d00fa0 | https://github.com/iamxpy/pointer_summarizer/tree/ebeb2ad32a45162c0da14dac0b6241b0b0d00fa0 | import math
import torch
import torch.nn as nn
import torch.nn.functional as F
class AffineLayer(nn.Module):
def __init__(self, dropout, d_model, d_ff):
super().__init__()
self.w_1 = nn.Linear(d_model, d_ff)
self.w_2 = nn.Linear(d_ff, d_model)
self.dropout = nn.Dropout(dropout)
def forward(self, x):
return self.w_2(self.dropout(F.relu(self.w_1(x))))
class MultiHeadedAttention(nn.Module):
def __init__(self, num_head, d_model, dropout=0.1):
super().__init__()
assert d_model % num_head == 0
self.d_k = d_model // num_head
self.h = num_head
self.linear_key = nn.Linear(d_model, d_model)
self.linear_value = nn.Linear(d_model, d_model)
self.linear_query = nn.Linear(d_model, d_model)
self.linear_out = nn.Linear(d_model, d_model)
self.dropout = nn.Dropout(p=dropout)
def attention(self, query, key, value, mask, dropout=None):
d_k = query.size(-1)
scores = torch.matmul(query, key.transpose(-2, -1)) / math.sqrt(d_k)
scores = scores.masked_fill(mask == 0, -1000000000.0)
p_attn = F.softmax(scores, dim=-1)
if dropout is not None:
p_attn = dropout(p_attn)
return torch.matmul(p_attn, value), p_attn
def forward(self, query, key, value, mask):
nbatches = query.size(0)
query = self.linear_query(query).view(nbatches, -1, self.h, self.d_k
).transpose(1, 2)
key = self.linear_key(key).view(nbatches, -1, self.h, self.d_k
).transpose(1, 2)
value = self.linear_value(value).view(nbatches, -1, self.h, self.d_k
).transpose(1, 2)
mask = mask.unsqueeze(1)
x, _attn = self.attention(query, key, value, mask, dropout=self.dropout
)
x = x.transpose(1, 2).contiguous().view(nbatches, -1, self.h * self.d_k
)
return self.linear_out(x)
class Model(nn.Module):
def __init__(self, num_head, dropout, d_model, d_ff):
super().__init__()
self.att_layer = MultiHeadedAttention(num_head, d_model, dropout)
self.norm_att = nn.LayerNorm(d_model)
self.dropout_att = nn.Dropout(dropout)
self.affine_layer = AffineLayer(dropout, d_model, d_ff)
self.norm_affine = nn.LayerNorm(d_model)
self.dropout_affine = nn.Dropout(dropout)
def forward(self, x, mask):
x_att = self.norm_att(x * mask)
x_att = self.att_layer(x_att, x_att, x_att, mask)
x = x + self.dropout_att(x_att)
x_affine = self.norm_affine(x * mask)
x_affine = self.affine_layer(x_affine)
return x + self.dropout_affine(x_affine)
def get_inputs():
return [torch.rand([4, 4, 4]), torch.rand([4, 4, 4])]
def get_init_inputs():
return [4, 0.5, 4, 4]
|
SoftDiceLossV2 | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/vc/cvcfr6g466rjwpqyuoei5b3fjc2q6re5nflgvwlz5ocdklbhptjy.py
# Topologically Sorted Source Nodes: [probs, mul, sum_1, add_1, sum_2], Original ATen: [aten.sigmoid, aten.mul, aten.sum, aten.add]
# Source node to ATen node mapping:
# add_1 => add_1
# mul => mul
# probs => sigmoid
# sum_1 => sum_1
# sum_2 => sum_2
# Graph fragment:
# %sigmoid : [num_users=2] = call_function[target=torch.ops.aten.sigmoid.default](args = (%arg0_1,), kwargs = {})
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sigmoid, %arg1_1), kwargs = {})
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%mul, [1, 2]), kwargs = {})
# %add_1 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%sigmoid, %arg1_1), kwargs = {})
# %sum_2 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%add_1, [1, 2]), kwargs = {})
triton_per_fused_add_mul_sigmoid_sum_0 = async_compile.triton('triton_per_fused_add_mul_sigmoid_sum_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[16, 16],
reduction_hint=ReductionHint.DEFAULT,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32', 5: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused_add_mul_sigmoid_sum_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 2, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused_add_mul_sigmoid_sum_0(in_ptr0, in_ptr1, out_ptr0, out_ptr1, xnumel, rnumel, XBLOCK : tl.constexpr):
xnumel = 16
rnumel = 16
RBLOCK: tl.constexpr = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
roffset = 0
rmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
r2 = rindex
x0 = xindex % 4
x1 = (xindex // 4)
x3 = xindex
tmp0 = tl.load(in_ptr0 + (x0 + (4*r2) + (64*x1)), xmask, other=0.0)
tmp2 = tl.load(in_ptr1 + (x0 + (4*r2) + (64*x1)), xmask, other=0.0)
tmp1 = tl.sigmoid(tmp0)
tmp3 = tmp1 * tmp2
tmp4 = tl.broadcast_to(tmp3, [XBLOCK, RBLOCK])
tmp6 = tl.where(xmask, tmp4, 0)
tmp7 = tl.sum(tmp6, 1)[:, None]
tmp8 = tmp1 + tmp2
tmp9 = tl.broadcast_to(tmp8, [XBLOCK, RBLOCK])
tmp11 = tl.where(xmask, tmp9, 0)
tmp12 = tl.sum(tmp11, 1)[:, None]
tl.store(out_ptr0 + (x3), tmp7, xmask)
tl.store(out_ptr1 + (x3), tmp12, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/2x/c2xcxojd4xdl4urjqd4gxqrmsugot4ql5kir7dkhanipimsb3jvz.py
# Topologically Sorted Source Nodes: [mul_1, numer, denor, truediv, loss, loss_1], Original ATen: [aten.mul, aten.add, aten.div, aten.rsub, aten.mean]
# Source node to ATen node mapping:
# denor => add_2
# loss => sub
# loss_1 => mean
# mul_1 => mul_1
# numer => add
# truediv => div
# Graph fragment:
# %mul_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sum_1, 2), kwargs = {})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul_1, 1), kwargs = {})
# %add_2 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%sum_2, 1), kwargs = {})
# %div : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%add, %add_2), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (1.0, %div), kwargs = {})
# %mean : [num_users=1] = call_function[target=torch.ops.aten.mean.default](args = (%sub,), kwargs = {})
triton_per_fused_add_div_mean_mul_rsub_1 = async_compile.triton('triton_per_fused_add_div_mean_mul_rsub_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[1, 16],
reduction_hint=ReductionHint.INNER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {3: 1}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 4), equal_to_1=(3,))]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused_add_div_mean_mul_rsub_1', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 1, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused_add_div_mean_mul_rsub_1(in_out_ptr0, in_ptr0, in_ptr1, xnumel, rnumel, XBLOCK : tl.constexpr):
xnumel = 1
rnumel = 16
RBLOCK: tl.constexpr = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[None, :]
roffset = 0
rmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
r0 = rindex
tmp0 = tl.load(in_ptr0 + (r0), None)
tmp5 = tl.load(in_ptr1 + (r0), None)
tmp1 = 2.0
tmp2 = tmp0 * tmp1
tmp3 = 1.0
tmp4 = tmp2 + tmp3
tmp6 = tmp5 + tmp3
tmp7 = tmp4 / tmp6
tmp8 = tmp3 - tmp7
tmp9 = tl.broadcast_to(tmp8, [XBLOCK, RBLOCK])
tmp11 = tl.sum(tmp9, 1)[:, None]
tmp12 = 16.0
tmp13 = tmp11 / tmp12
tl.debug_barrier()
tl.store(in_out_ptr0 + (tl.full([XBLOCK, 1], 0, tl.int32)), tmp13, None)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, arg1_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg1_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
buf1 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [probs, mul, sum_1, add_1, sum_2], Original ATen: [aten.sigmoid, aten.mul, aten.sum, aten.add]
stream0 = get_raw_stream(0)
triton_per_fused_add_mul_sigmoid_sum_0.run(arg0_1, arg1_1, buf0, buf1, 16, 16, grid=grid(16), stream=stream0)
del arg0_1
del arg1_1
buf2 = empty_strided_cuda((), (), torch.float32)
buf3 = buf2; del buf2 # reuse
# Topologically Sorted Source Nodes: [mul_1, numer, denor, truediv, loss, loss_1], Original ATen: [aten.mul, aten.add, aten.div, aten.rsub, aten.mean]
triton_per_fused_add_div_mean_mul_rsub_1.run(buf3, buf0, buf1, 1, 16, grid=grid(1), stream=stream0)
del buf0
del buf1
return (buf3, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
arg1_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([arg0_1, arg1_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.cuda.amp as amp
class SoftDiceLossV2Func(torch.autograd.Function):
"""
compute backward directly for better numeric stability
"""
@staticmethod
@amp.custom_fwd
def forward(ctx, logits, labels, p, smooth):
logits = logits.float()
probs = torch.sigmoid(logits)
numer = 2 * (probs * labels).sum(dim=(1, 2)) + smooth
denor = (probs.pow(p) + labels).sum(dim=(1, 2)) + smooth
loss = 1.0 - numer / denor
ctx.vars = probs, labels, numer, denor, p, smooth
return loss
@staticmethod
@amp.custom_bwd
def backward(ctx, grad_output):
"""
compute gradient of soft-dice loss
"""
probs, labels, numer, denor, p, _smooth = ctx.vars
M = numer.view(-1, 1, 1) - (probs * labels).mul_(2)
N = denor.view(-1, 1, 1) - probs.pow(p)
mppi_1 = probs.pow(p - 1).mul_(p).mul_(M)
grads = torch.where(labels == 1, probs.pow(p).mul_(2 * (1.0 - p)) -
mppi_1 + N.mul_(2), -mppi_1)
grads = grads.div_((probs.pow(p) + N).pow(2)).mul_(probs).mul_(1.0 -
probs)
grads = grads.mul_(grad_output.view(-1, 1, 1)).neg_()
return grads, None, None, None
class SoftDiceLossV2(nn.Module):
"""
soft-dice loss, useful in binary segmentation
"""
def __init__(self, p=1, smooth=1, reduction='mean'):
super(SoftDiceLossV2, self).__init__()
self.p = p
self.smooth = smooth
self.reduction = reduction
def forward(self, logits, labels):
"""
args: logits: tensor of shape (N, H, W)
args: label: tensor of shape(N, H, W)
"""
loss = SoftDiceLossV2Func.apply(logits, labels, self.p, self.smooth)
if self.reduction == 'mean':
loss = loss.mean()
elif self.reduction == 'sum':
loss = loss.sum()
return loss
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {}]
| import torch
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
import torch.nn as nn
import torch.cuda.amp as amp
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_per_fused_add_mul_sigmoid_sum_0(in_ptr0, in_ptr1, out_ptr0,
out_ptr1, xnumel, rnumel, XBLOCK: tl.constexpr):
xnumel = 16
RBLOCK: tl.constexpr = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
r2 = rindex
x0 = xindex % 4
x1 = xindex // 4
x3 = xindex
tmp0 = tl.load(in_ptr0 + (x0 + 4 * r2 + 64 * x1), xmask, other=0.0)
tmp2 = tl.load(in_ptr1 + (x0 + 4 * r2 + 64 * x1), xmask, other=0.0)
tmp1 = tl.sigmoid(tmp0)
tmp3 = tmp1 * tmp2
tmp4 = tl.broadcast_to(tmp3, [XBLOCK, RBLOCK])
tmp6 = tl.where(xmask, tmp4, 0)
tmp7 = tl.sum(tmp6, 1)[:, None]
tmp8 = tmp1 + tmp2
tmp9 = tl.broadcast_to(tmp8, [XBLOCK, RBLOCK])
tmp11 = tl.where(xmask, tmp9, 0)
tmp12 = tl.sum(tmp11, 1)[:, None]
tl.store(out_ptr0 + x3, tmp7, xmask)
tl.store(out_ptr1 + x3, tmp12, xmask)
@triton.jit
def triton_per_fused_add_div_mean_mul_rsub_1(in_out_ptr0, in_ptr0, in_ptr1,
xnumel, rnumel, XBLOCK: tl.constexpr):
RBLOCK: tl.constexpr = 16
xoffset = tl.program_id(0) * XBLOCK
xoffset + tl.arange(0, XBLOCK)[:, None]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[None, :]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
r0 = rindex
tmp0 = tl.load(in_ptr0 + r0, None)
tmp5 = tl.load(in_ptr1 + r0, None)
tmp1 = 2.0
tmp2 = tmp0 * tmp1
tmp3 = 1.0
tmp4 = tmp2 + tmp3
tmp6 = tmp5 + tmp3
tmp7 = tmp4 / tmp6
tmp8 = tmp3 - tmp7
tmp9 = tl.broadcast_to(tmp8, [XBLOCK, RBLOCK])
tmp11 = tl.sum(tmp9, 1)[:, None]
tmp12 = 16.0
tmp13 = tmp11 / tmp12
tl.debug_barrier()
tl.store(in_out_ptr0 + tl.full([XBLOCK, 1], 0, tl.int32), tmp13, None)
def call(args):
arg0_1, arg1_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg1_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
buf1 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
get_raw_stream(0)
triton_per_fused_add_mul_sigmoid_sum_0[grid(16)](arg0_1, arg1_1,
buf0, buf1, 16, 16, XBLOCK=1, num_warps=2, num_stages=1)
del arg0_1
del arg1_1
buf2 = empty_strided_cuda((), (), torch.float32)
buf3 = buf2
del buf2
triton_per_fused_add_div_mean_mul_rsub_1[grid(1)](buf3, buf0, buf1,
1, 16, XBLOCK=1, num_warps=2, num_stages=1)
del buf0
del buf1
return buf3,
class SoftDiceLossV2Func(torch.autograd.Function):
"""
compute backward directly for better numeric stability
"""
@staticmethod
@amp.custom_fwd
def forward(ctx, logits, labels, p, smooth):
logits = logits.float()
probs = torch.sigmoid(logits)
numer = 2 * (probs * labels).sum(dim=(1, 2)) + smooth
denor = (probs.pow(p) + labels).sum(dim=(1, 2)) + smooth
loss = 1.0 - numer / denor
ctx.vars = probs, labels, numer, denor, p, smooth
return loss
@staticmethod
@amp.custom_bwd
def backward(ctx, grad_output):
"""
compute gradient of soft-dice loss
"""
probs, labels, numer, denor, p, _smooth = ctx.vars
M = numer.view(-1, 1, 1) - (probs * labels).mul_(2)
N = denor.view(-1, 1, 1) - probs.pow(p)
mppi_1 = probs.pow(p - 1).mul_(p).mul_(M)
grads = torch.where(labels == 1, probs.pow(p).mul_(2 * (1.0 - p)) -
mppi_1 + N.mul_(2), -mppi_1)
grads = grads.div_((probs.pow(p) + N).pow(2)).mul_(probs).mul_(1.0 -
probs)
grads = grads.mul_(grad_output.view(-1, 1, 1)).neg_()
return grads, None, None, None
class SoftDiceLossV2New(nn.Module):
"""
soft-dice loss, useful in binary segmentation
"""
def __init__(self, p=1, smooth=1, reduction='mean'):
super(SoftDiceLossV2New, self).__init__()
self.p = p
self.smooth = smooth
self.reduction = reduction
def forward(self, input_0, input_1):
arg0_1 = input_0
arg1_1 = input_1
output = call([arg0_1, arg1_1])
return output[0]
| imvladikon/pytorch-loss | SoftDiceLossV2 | false | 6,879 | [
"MIT"
] | 1 | 6cfaabe1be898e1ff000b3dffb46d0ef09096f6b | https://github.com/imvladikon/pytorch-loss/tree/6cfaabe1be898e1ff000b3dffb46d0ef09096f6b | import torch
import torch.nn as nn
import torch.cuda.amp as amp
class SoftDiceLossV2Func(torch.autograd.Function):
"""
compute backward directly for better numeric stability
"""
@staticmethod
@amp.custom_fwd
def forward(ctx, logits, labels, p, smooth):
logits = logits.float()
probs = torch.sigmoid(logits)
numer = 2 * (probs * labels).sum(dim=(1, 2)) + smooth
denor = (probs.pow(p) + labels).sum(dim=(1, 2)) + smooth
loss = 1.0 - numer / denor
ctx.vars = probs, labels, numer, denor, p, smooth
return loss
@staticmethod
@amp.custom_bwd
def backward(ctx, grad_output):
"""
compute gradient of soft-dice loss
"""
probs, labels, numer, denor, p, _smooth = ctx.vars
M = numer.view(-1, 1, 1) - (probs * labels).mul_(2)
N = denor.view(-1, 1, 1) - probs.pow(p)
mppi_1 = probs.pow(p - 1).mul_(p).mul_(M)
grads = torch.where(labels == 1, probs.pow(p).mul_(2 * (1.0 - p)) -
mppi_1 + N.mul_(2), -mppi_1)
grads = grads.div_((probs.pow(p) + N).pow(2)).mul_(probs).mul_(1.0 -
probs)
grads = grads.mul_(grad_output.view(-1, 1, 1)).neg_()
return grads, None, None, None
class Model(nn.Module):
"""
soft-dice loss, useful in binary segmentation
"""
def __init__(self, p=1, smooth=1, reduction='mean'):
super().__init__()
self.p = p
self.smooth = smooth
self.reduction = reduction
def forward(self, logits, labels):
"""
args: logits: tensor of shape (N, H, W)
args: label: tensor of shape(N, H, W)
"""
loss = SoftDiceLossV2Func.apply(logits, labels, self.p, self.smooth)
if self.reduction == 'mean':
loss = loss.mean()
elif self.reduction == 'sum':
loss = loss.sum()
return loss
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return []
|
DilConv1dWithGLU | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/sr/csrzlteph4svc746shxzwrfzfygp3ngujwxcrnvcusqhc43dtftf.py
# Topologically Sorted Source Nodes: [layer_norm], Original ATen: [aten.native_layer_norm]
# Source node to ATen node mapping:
# layer_norm => add, clone, rsqrt, var_mean
# Graph fragment:
# %clone : [num_users=2] = call_function[target=torch.ops.aten.clone.default](args = (%permute,), kwargs = {memory_format: torch.contiguous_format})
# %var_mean : [num_users=2] = call_function[target=torch.ops.aten.var_mean.correction](args = (%clone, [2]), kwargs = {correction: 0, keepdim: True})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%getitem, 1e-05), kwargs = {})
# %rsqrt : [num_users=1] = call_function[target=torch.ops.aten.rsqrt.default](args = (%add,), kwargs = {})
triton_poi_fused_native_layer_norm_0 = async_compile.triton('triton_poi_fused_native_layer_norm_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_native_layer_norm_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 4, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_native_layer_norm_0(in_ptr0, out_ptr0, out_ptr1, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 4
x1 = (xindex // 4)
x2 = xindex
tmp0 = tl.load(in_ptr0 + (x0 + (16*x1)), xmask)
tmp1 = tl.load(in_ptr0 + (4 + x0 + (16*x1)), xmask)
tmp3 = tl.load(in_ptr0 + (8 + x0 + (16*x1)), xmask)
tmp5 = tl.load(in_ptr0 + (12 + x0 + (16*x1)), xmask)
tmp2 = tmp0 + tmp1
tmp4 = tmp2 + tmp3
tmp6 = tmp4 + tmp5
tmp7 = 4.0
tmp8 = tmp6 / tmp7
tmp9 = tmp0 - tmp8
tmp10 = tmp9 * tmp9
tmp11 = tmp1 - tmp8
tmp12 = tmp11 * tmp11
tmp13 = tmp10 + tmp12
tmp14 = tmp3 - tmp8
tmp15 = tmp14 * tmp14
tmp16 = tmp13 + tmp15
tmp17 = tmp5 - tmp8
tmp18 = tmp17 * tmp17
tmp19 = tmp16 + tmp18
tmp20 = tmp19 / tmp7
tmp21 = 1e-05
tmp22 = tmp20 + tmp21
tmp23 = libdevice.rsqrt(tmp22)
tl.store(out_ptr0 + (x2), tmp8, xmask)
tl.store(out_ptr1 + (x2), tmp23, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/ep/cepzni4orknhaka6ezxewg2ehebbjiupi52qelfw47wqs4ekerxq.py
# Topologically Sorted Source Nodes: [layer_norm, x_1, x_2], Original ATen: [aten.native_layer_norm, aten.leaky_relu, aten.convolution]
# Source node to ATen node mapping:
# layer_norm => add, add_1, clone, mul, mul_1, rsqrt, sub, var_mean
# x_1 => gt, mul_2, where
# x_2 => convolution
# Graph fragment:
# %clone : [num_users=2] = call_function[target=torch.ops.aten.clone.default](args = (%permute,), kwargs = {memory_format: torch.contiguous_format})
# %var_mean : [num_users=2] = call_function[target=torch.ops.aten.var_mean.correction](args = (%clone, [2]), kwargs = {correction: 0, keepdim: True})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%getitem, 1e-05), kwargs = {})
# %rsqrt : [num_users=1] = call_function[target=torch.ops.aten.rsqrt.default](args = (%add,), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%clone, %getitem_1), kwargs = {})
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sub, %rsqrt), kwargs = {})
# %mul_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%mul, %primals_2), kwargs = {})
# %add_1 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul_1, %primals_3), kwargs = {})
# %gt : [num_users=1] = call_function[target=torch.ops.aten.gt.Scalar](args = (%permute_1, 0), kwargs = {})
# %mul_2 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%permute_1, 0.01), kwargs = {})
# %where : [num_users=2] = call_function[target=torch.ops.aten.where.self](args = (%gt, %permute_1, %mul_2), kwargs = {})
# %convolution : [num_users=2] = call_function[target=torch.ops.aten.convolution.default](args = (%where, %primals_4, %primals_5, [1], [0], [1], False, [0], 1), kwargs = {})
triton_poi_fused_convolution_leaky_relu_native_layer_norm_1 = async_compile.triton('triton_poi_fused_convolution_leaky_relu_native_layer_norm_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16, 4], tile_hint=TileHint.DEFAULT,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: '*fp32', 5: '*fp32', 6: '*fp32', 7: 'i32', 8: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5, 6, 7), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_leaky_relu_native_layer_norm_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_leaky_relu_native_layer_norm_1(in_ptr0, in_ptr1, in_ptr2, in_ptr3, in_ptr4, out_ptr1, out_ptr2, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y3 = yindex
y1 = (yindex // 4)
y0 = yindex % 4
tmp0 = tl.load(in_ptr0 + (x2 + (4*y3)), xmask & ymask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (x2 + (4*y1)), xmask & ymask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr2 + (x2 + (4*y1)), xmask & ymask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr3 + (y0), ymask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr4 + (y0), ymask, eviction_policy='evict_last')
tmp2 = tmp0 - tmp1
tmp4 = tmp2 * tmp3
tmp6 = tmp4 * tmp5
tmp8 = tmp6 + tmp7
tmp9 = 0.0
tmp10 = tmp8 > tmp9
tmp11 = 0.01
tmp12 = tmp8 * tmp11
tmp13 = tl.where(tmp10, tmp8, tmp12)
tl.store(out_ptr1 + (y0 + (4*x2) + (16*y1)), tmp13, xmask & ymask)
tl.store(out_ptr2 + (x2 + (4*y3)), tmp13, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/lf/clf7hs52i4bd5d3e73uio27ntyjfqmszkbsw6dta3r6rzgeftva3.py
# Topologically Sorted Source Nodes: [x_2], Original ATen: [aten.convolution]
# Source node to ATen node mapping:
# x_2 => convolution
# Graph fragment:
# %convolution : [num_users=2] = call_function[target=torch.ops.aten.convolution.default](args = (%where, %primals_4, %primals_5, [1], [0], [1], False, [0], 1), kwargs = {})
triton_poi_fused_convolution_2 = async_compile.triton('triton_poi_fused_convolution_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_2', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_2(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = (xindex // 4) % 4
tmp0 = tl.load(in_out_ptr0 + (x3), xmask)
tmp1 = tl.load(in_ptr0 + (x1), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + (x3), tmp2, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/2f/c2fmbms5yu7pndqccuviyfk3uh5ctm4s4m5duyocimrn6uqrpnx3.py
# Topologically Sorted Source Nodes: [conv1d_1], Original ATen: [aten.convolution]
# Source node to ATen node mapping:
# conv1d_1 => convolution_1
# Graph fragment:
# %convolution_1 : [num_users=2] = call_function[target=torch.ops.aten.convolution.default](args = (%where_1, %primals_8, %primals_9, [1], [1], [1], False, [0], 1), kwargs = {})
triton_poi_fused_convolution_3 = async_compile.triton('triton_poi_fused_convolution_3', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[128],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_3', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_3(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 80
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = (xindex // 5) % 4
tmp0 = tl.load(in_out_ptr0 + (x3), xmask)
tmp1 = tl.load(in_ptr0 + (x1), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + (x3), tmp2, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/6m/c6mzkhocn3islr7c5jwkrgvxbykm5jywftruw7si2p7sv5dljcno.py
# Topologically Sorted Source Nodes: [layer_norm_2], Original ATen: [aten.native_layer_norm]
# Source node to ATen node mapping:
# layer_norm_2 => add_4, clone_2, rsqrt_2, var_mean_2
# Graph fragment:
# %clone_2 : [num_users=2] = call_function[target=torch.ops.aten.clone.default](args = (%permute_4,), kwargs = {memory_format: torch.contiguous_format})
# %var_mean_2 : [num_users=2] = call_function[target=torch.ops.aten.var_mean.correction](args = (%clone_2, [2]), kwargs = {correction: 0, keepdim: True})
# %add_4 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%getitem_4, 1e-05), kwargs = {})
# %rsqrt_2 : [num_users=1] = call_function[target=torch.ops.aten.rsqrt.default](args = (%add_4,), kwargs = {})
triton_poi_fused_native_layer_norm_4 = async_compile.triton('triton_poi_fused_native_layer_norm_4', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_native_layer_norm_4', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 4, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_native_layer_norm_4(in_ptr0, out_ptr0, out_ptr1, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 4
x1 = (xindex // 4)
x2 = xindex
tmp0 = tl.load(in_ptr0 + (x0 + (20*x1)), xmask)
tmp1 = tl.load(in_ptr0 + (5 + x0 + (20*x1)), xmask)
tmp3 = tl.load(in_ptr0 + (10 + x0 + (20*x1)), xmask)
tmp5 = tl.load(in_ptr0 + (15 + x0 + (20*x1)), xmask)
tmp2 = tmp0 + tmp1
tmp4 = tmp2 + tmp3
tmp6 = tmp4 + tmp5
tmp7 = 4.0
tmp8 = tmp6 / tmp7
tmp9 = tmp0 - tmp8
tmp10 = tmp9 * tmp9
tmp11 = tmp1 - tmp8
tmp12 = tmp11 * tmp11
tmp13 = tmp10 + tmp12
tmp14 = tmp3 - tmp8
tmp15 = tmp14 * tmp14
tmp16 = tmp13 + tmp15
tmp17 = tmp5 - tmp8
tmp18 = tmp17 * tmp17
tmp19 = tmp16 + tmp18
tmp20 = tmp19 / tmp7
tmp21 = 1e-05
tmp22 = tmp20 + tmp21
tmp23 = libdevice.rsqrt(tmp22)
tl.store(out_ptr0 + (x2), tmp8, xmask)
tl.store(out_ptr1 + (x2), tmp23, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/nu/cnuopqsoegbagggwrkx7mrklpmg45rhcdsbpd5bwecwtnbzabt6t.py
# Topologically Sorted Source Nodes: [layer_norm_2, x_7, conv1d_2, conv1d_3], Original ATen: [aten.native_layer_norm, aten.leaky_relu, aten.convolution]
# Source node to ATen node mapping:
# conv1d_2 => convolution_2
# conv1d_3 => convolution_3
# layer_norm_2 => add_4, add_5, clone_2, mul_6, mul_7, rsqrt_2, sub_2, var_mean_2
# x_7 => gt_2, mul_8, where_2
# Graph fragment:
# %clone_2 : [num_users=2] = call_function[target=torch.ops.aten.clone.default](args = (%permute_4,), kwargs = {memory_format: torch.contiguous_format})
# %var_mean_2 : [num_users=2] = call_function[target=torch.ops.aten.var_mean.correction](args = (%clone_2, [2]), kwargs = {correction: 0, keepdim: True})
# %add_4 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%getitem_4, 1e-05), kwargs = {})
# %rsqrt_2 : [num_users=1] = call_function[target=torch.ops.aten.rsqrt.default](args = (%add_4,), kwargs = {})
# %sub_2 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%clone_2, %getitem_5), kwargs = {})
# %mul_6 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sub_2, %rsqrt_2), kwargs = {})
# %mul_7 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%mul_6, %primals_10), kwargs = {})
# %add_5 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul_7, %primals_11), kwargs = {})
# %gt_2 : [num_users=1] = call_function[target=torch.ops.aten.gt.Scalar](args = (%permute_5, 0), kwargs = {})
# %mul_8 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%permute_5, 0.01), kwargs = {})
# %where_2 : [num_users=3] = call_function[target=torch.ops.aten.where.self](args = (%gt_2, %permute_5, %mul_8), kwargs = {})
# %convolution_2 : [num_users=2] = call_function[target=torch.ops.aten.convolution.default](args = (%where_2, %primals_12, %primals_13, [1], [0], [1], False, [0], 1), kwargs = {})
# %convolution_3 : [num_users=2] = call_function[target=torch.ops.aten.convolution.default](args = (%where_2, %primals_14, %primals_15, [1], [0], [1], False, [0], 1), kwargs = {})
triton_poi_fused_convolution_leaky_relu_native_layer_norm_5 = async_compile.triton('triton_poi_fused_convolution_leaky_relu_native_layer_norm_5', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16, 4], tile_hint=TileHint.DEFAULT,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: '*fp32', 5: '*fp32', 6: '*fp32', 7: '*fp32', 8: 'i32', 9: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5, 6, 7, 8), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_leaky_relu_native_layer_norm_5', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_leaky_relu_native_layer_norm_5(in_ptr0, in_ptr1, in_ptr2, in_ptr3, in_ptr4, out_ptr1, out_ptr2, out_ptr3, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y3 = yindex
y1 = (yindex // 4)
y0 = yindex % 4
tmp0 = tl.load(in_ptr0 + (x2 + (5*y3)), xmask & ymask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (x2 + (4*y1)), xmask & ymask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr2 + (x2 + (4*y1)), xmask & ymask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr3 + (y0), ymask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr4 + (y0), ymask, eviction_policy='evict_last')
tmp2 = tmp0 - tmp1
tmp4 = tmp2 * tmp3
tmp6 = tmp4 * tmp5
tmp8 = tmp6 + tmp7
tmp9 = 0.0
tmp10 = tmp8 > tmp9
tmp11 = 0.01
tmp12 = tmp8 * tmp11
tmp13 = tl.where(tmp10, tmp8, tmp12)
tl.store(out_ptr1 + (y0 + (4*x2) + (16*y1)), tmp13, xmask & ymask)
tl.store(out_ptr2 + (x2 + (4*y3)), tmp13, xmask & ymask)
tl.store(out_ptr3 + (x2 + (4*y3)), tmp13, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/5x/c5xlvcqikyssfno5q53dmxs4ui2xjosb4nsvzxrcridsmdua2jff.py
# Topologically Sorted Source Nodes: [conv1d_2, conv1d_3, sigmoid, x_8, x_9], Original ATen: [aten.convolution, aten.sigmoid, aten.mul, aten.add]
# Source node to ATen node mapping:
# conv1d_2 => convolution_2
# conv1d_3 => convolution_3
# sigmoid => sigmoid
# x_8 => mul_9
# x_9 => add_6
# Graph fragment:
# %convolution_2 : [num_users=2] = call_function[target=torch.ops.aten.convolution.default](args = (%where_2, %primals_12, %primals_13, [1], [0], [1], False, [0], 1), kwargs = {})
# %convolution_3 : [num_users=2] = call_function[target=torch.ops.aten.convolution.default](args = (%where_2, %primals_14, %primals_15, [1], [0], [1], False, [0], 1), kwargs = {})
# %sigmoid : [num_users=1] = call_function[target=torch.ops.aten.sigmoid.default](args = (%convolution_3,), kwargs = {})
# %mul_9 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%convolution_2, %sigmoid), kwargs = {})
# %add_6 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul_9, %primals_1), kwargs = {})
triton_poi_fused_add_convolution_mul_sigmoid_6 = async_compile.triton('triton_poi_fused_add_convolution_mul_sigmoid_6', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: '*fp32', 5: '*fp32', 6: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5, 6), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_convolution_mul_sigmoid_6', 'mutated_arg_names': ['in_out_ptr0', 'in_out_ptr1'], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_convolution_mul_sigmoid_6(in_out_ptr0, in_out_ptr1, in_ptr0, in_ptr1, in_ptr2, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = (xindex // 4) % 4
tmp0 = tl.load(in_out_ptr0 + (x3), xmask)
tmp1 = tl.load(in_ptr0 + (x1), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_out_ptr1 + (x3), xmask)
tmp4 = tl.load(in_ptr1 + (x1), xmask, eviction_policy='evict_last')
tmp8 = tl.load(in_ptr2 + (x3), xmask)
tmp2 = tmp0 + tmp1
tmp5 = tmp3 + tmp4
tmp6 = tl.sigmoid(tmp5)
tmp7 = tmp2 * tmp6
tmp9 = tmp7 + tmp8
tl.store(in_out_ptr0 + (x3), tmp2, xmask)
tl.store(in_out_ptr1 + (x3), tmp5, xmask)
tl.store(out_ptr0 + (x3), tmp9, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9, primals_10, primals_11, primals_12, primals_13, primals_14, primals_15 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_2, (4, ), (1, ))
assert_size_stride(primals_3, (4, ), (1, ))
assert_size_stride(primals_4, (4, 4, 1), (4, 1, 1))
assert_size_stride(primals_5, (4, ), (1, ))
assert_size_stride(primals_6, (4, ), (1, ))
assert_size_stride(primals_7, (4, ), (1, ))
assert_size_stride(primals_8, (4, 4, 2), (8, 2, 1))
assert_size_stride(primals_9, (4, ), (1, ))
assert_size_stride(primals_10, (4, ), (1, ))
assert_size_stride(primals_11, (4, ), (1, ))
assert_size_stride(primals_12, (4, 4, 1), (4, 1, 1))
assert_size_stride(primals_13, (4, ), (1, ))
assert_size_stride(primals_14, (4, 4, 1), (4, 1, 1))
assert_size_stride(primals_15, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 1), (4, 1, 16), torch.float32)
buf1 = empty_strided_cuda((4, 4, 1), (4, 1, 16), torch.float32)
# Topologically Sorted Source Nodes: [layer_norm], Original ATen: [aten.native_layer_norm]
stream0 = get_raw_stream(0)
triton_poi_fused_native_layer_norm_0.run(primals_1, buf0, buf1, 16, grid=grid(16), stream=stream0)
buf3 = empty_strided_cuda((4, 4, 4), (16, 1, 4), torch.float32)
buf4 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [layer_norm, x_1, x_2], Original ATen: [aten.native_layer_norm, aten.leaky_relu, aten.convolution]
triton_poi_fused_convolution_leaky_relu_native_layer_norm_1.run(primals_1, buf0, buf1, primals_2, primals_3, buf3, buf4, 16, 4, grid=grid(16, 4), stream=stream0)
# Topologically Sorted Source Nodes: [x_2], Original ATen: [aten.convolution]
buf5 = extern_kernels.convolution(buf4, primals_4, stride=(1,), padding=(0,), dilation=(1,), transposed=False, output_padding=(0,), groups=1, bias=None)
assert_size_stride(buf5, (4, 4, 4), (16, 4, 1))
buf6 = buf5; del buf5 # reuse
# Topologically Sorted Source Nodes: [x_2], Original ATen: [aten.convolution]
triton_poi_fused_convolution_2.run(buf6, primals_5, 64, grid=grid(64), stream=stream0)
del primals_5
buf7 = buf1; del buf1 # reuse
buf8 = buf0; del buf0 # reuse
# Topologically Sorted Source Nodes: [layer_norm_1], Original ATen: [aten.native_layer_norm]
triton_poi_fused_native_layer_norm_0.run(buf6, buf7, buf8, 16, grid=grid(16), stream=stream0)
buf10 = reinterpret_tensor(buf4, (4, 4, 4), (16, 1, 4), 0); del buf4 # reuse
buf11 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [layer_norm_1, x_4, conv1d_1], Original ATen: [aten.native_layer_norm, aten.leaky_relu, aten.convolution]
triton_poi_fused_convolution_leaky_relu_native_layer_norm_1.run(buf6, buf7, buf8, primals_6, primals_7, buf10, buf11, 16, 4, grid=grid(16, 4), stream=stream0)
# Topologically Sorted Source Nodes: [conv1d_1], Original ATen: [aten.convolution]
buf12 = extern_kernels.convolution(buf11, primals_8, stride=(1,), padding=(1,), dilation=(1,), transposed=False, output_padding=(0,), groups=1, bias=None)
assert_size_stride(buf12, (4, 4, 5), (20, 5, 1))
buf13 = buf12; del buf12 # reuse
# Topologically Sorted Source Nodes: [conv1d_1], Original ATen: [aten.convolution]
triton_poi_fused_convolution_3.run(buf13, primals_9, 80, grid=grid(80), stream=stream0)
del primals_9
buf14 = buf8; del buf8 # reuse
buf15 = buf7; del buf7 # reuse
# Topologically Sorted Source Nodes: [layer_norm_2], Original ATen: [aten.native_layer_norm]
triton_poi_fused_native_layer_norm_4.run(buf13, buf14, buf15, 16, grid=grid(16), stream=stream0)
buf17 = reinterpret_tensor(buf11, (4, 4, 4), (16, 1, 4), 0); del buf11 # reuse
buf18 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
buf21 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [layer_norm_2, x_7, conv1d_2, conv1d_3], Original ATen: [aten.native_layer_norm, aten.leaky_relu, aten.convolution]
triton_poi_fused_convolution_leaky_relu_native_layer_norm_5.run(buf13, buf14, buf15, primals_10, primals_11, buf17, buf18, buf21, 16, 4, grid=grid(16, 4), stream=stream0)
del buf14
del buf15
# Topologically Sorted Source Nodes: [conv1d_2], Original ATen: [aten.convolution]
buf19 = extern_kernels.convolution(buf18, primals_12, stride=(1,), padding=(0,), dilation=(1,), transposed=False, output_padding=(0,), groups=1, bias=None)
assert_size_stride(buf19, (4, 4, 4), (16, 4, 1))
del buf18
# Topologically Sorted Source Nodes: [conv1d_3], Original ATen: [aten.convolution]
buf22 = extern_kernels.convolution(buf21, primals_14, stride=(1,), padding=(0,), dilation=(1,), transposed=False, output_padding=(0,), groups=1, bias=None)
assert_size_stride(buf22, (4, 4, 4), (16, 4, 1))
buf20 = buf19; del buf19 # reuse
buf23 = buf22; del buf22 # reuse
buf24 = buf21; del buf21 # reuse
# Topologically Sorted Source Nodes: [conv1d_2, conv1d_3, sigmoid, x_8, x_9], Original ATen: [aten.convolution, aten.sigmoid, aten.mul, aten.add]
triton_poi_fused_add_convolution_mul_sigmoid_6.run(buf20, buf23, primals_13, primals_15, primals_1, buf24, 64, grid=grid(64), stream=stream0)
del primals_13
del primals_15
return (buf24, primals_1, primals_2, primals_3, primals_4, primals_6, primals_7, primals_8, primals_10, primals_11, primals_12, primals_14, buf3, buf6, buf10, buf13, buf17, buf20, buf23, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, 4, 1), (4, 1, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_8 = rand_strided((4, 4, 2), (8, 2, 1), device='cuda:0', dtype=torch.float32)
primals_9 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_10 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_11 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_12 = rand_strided((4, 4, 1), (4, 1, 1), device='cuda:0', dtype=torch.float32)
primals_13 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_14 = rand_strided((4, 4, 1), (4, 1, 1), device='cuda:0', dtype=torch.float32)
primals_15 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9, primals_10, primals_11, primals_12, primals_13, primals_14, primals_15])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.nn.functional as F
class DilConv1dWithGLU(nn.Module):
def __init__(self, num_channels, dilation, lenght=100, kernel_size=2,
activation=F.leaky_relu, residual_connection=True, dropout=0.2):
super(DilConv1dWithGLU, self).__init__()
self.dilation = dilation
self.start_ln = nn.LayerNorm(num_channels)
self.start_conv1x1 = nn.Conv1d(num_channels, num_channels,
kernel_size=1)
self.dilconv_ln = nn.LayerNorm(num_channels)
self.dilated_conv = nn.Conv1d(num_channels, num_channels, dilation=
dilation, kernel_size=kernel_size, padding=dilation)
self.gate_ln = nn.LayerNorm(num_channels)
self.end_conv1x1 = nn.Conv1d(num_channels, num_channels, kernel_size=1)
self.gated_conv1x1 = nn.Conv1d(num_channels, num_channels,
kernel_size=1)
self.activation = activation
self.buffer = None
self.residual_connection = residual_connection
def clear_buffer(self):
self.buffer = None
def forward(self, x_inp, sampling=False):
x = self.start_ln(x_inp.transpose(1, 2)).transpose(1, 2)
x = self.activation(x)
x = self.start_conv1x1(x)
x = self.dilconv_ln(x.transpose(1, 2)).transpose(1, 2)
x = self.activation(x)
if sampling:
if self.buffer is None:
self.buffer = x
else:
pre_buffer = torch.cat([self.buffer, x], dim=2)
self.buffer = pre_buffer[:, :, -(self.dilation + 1):]
if self.buffer.shape[2] == self.dilation + 1:
x = self.buffer
else:
x = torch.cat([torch.zeros(self.buffer.shape[0], self.
buffer.shape[1], self.dilation + 1 - self.buffer.shape[
2], device=x_inp.device), self.buffer], dim=2)
x = self.dilated_conv(x)[:, :, self.dilation:]
x = x[:, :, :x_inp.shape[-1]]
else:
x = self.dilated_conv(x)[:, :, :x_inp.shape[-1]]
x = self.gate_ln(x.transpose(1, 2)).transpose(1, 2)
x = self.activation(x)
x = self.end_conv1x1(x) * torch.sigmoid(self.gated_conv1x1(x))
if self.residual_connection:
x = x + x_inp
return x
def get_inputs():
return [torch.rand([4, 4, 4])]
def get_init_inputs():
return [[], {'num_channels': 4, 'dilation': 1}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime.triton_helpers import libdevice
import torch.nn as nn
import torch.nn.functional as F
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_native_layer_norm_0(in_ptr0, out_ptr0, out_ptr1,
xnumel, XBLOCK: tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 4
x1 = xindex // 4
x2 = xindex
tmp0 = tl.load(in_ptr0 + (x0 + 16 * x1), xmask)
tmp1 = tl.load(in_ptr0 + (4 + x0 + 16 * x1), xmask)
tmp3 = tl.load(in_ptr0 + (8 + x0 + 16 * x1), xmask)
tmp5 = tl.load(in_ptr0 + (12 + x0 + 16 * x1), xmask)
tmp2 = tmp0 + tmp1
tmp4 = tmp2 + tmp3
tmp6 = tmp4 + tmp5
tmp7 = 4.0
tmp8 = tmp6 / tmp7
tmp9 = tmp0 - tmp8
tmp10 = tmp9 * tmp9
tmp11 = tmp1 - tmp8
tmp12 = tmp11 * tmp11
tmp13 = tmp10 + tmp12
tmp14 = tmp3 - tmp8
tmp15 = tmp14 * tmp14
tmp16 = tmp13 + tmp15
tmp17 = tmp5 - tmp8
tmp18 = tmp17 * tmp17
tmp19 = tmp16 + tmp18
tmp20 = tmp19 / tmp7
tmp21 = 1e-05
tmp22 = tmp20 + tmp21
tmp23 = libdevice.rsqrt(tmp22)
tl.store(out_ptr0 + x2, tmp8, xmask)
tl.store(out_ptr1 + x2, tmp23, xmask)
@triton.jit
def triton_poi_fused_convolution_leaky_relu_native_layer_norm_1(in_ptr0,
in_ptr1, in_ptr2, in_ptr3, in_ptr4, out_ptr1, out_ptr2, ynumel, xnumel,
YBLOCK: tl.constexpr, XBLOCK: tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y3 = yindex
y1 = yindex // 4
y0 = yindex % 4
tmp0 = tl.load(in_ptr0 + (x2 + 4 * y3), xmask & ymask, eviction_policy=
'evict_last')
tmp1 = tl.load(in_ptr1 + (x2 + 4 * y1), xmask & ymask, eviction_policy=
'evict_last')
tmp3 = tl.load(in_ptr2 + (x2 + 4 * y1), xmask & ymask, eviction_policy=
'evict_last')
tmp5 = tl.load(in_ptr3 + y0, ymask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr4 + y0, ymask, eviction_policy='evict_last')
tmp2 = tmp0 - tmp1
tmp4 = tmp2 * tmp3
tmp6 = tmp4 * tmp5
tmp8 = tmp6 + tmp7
tmp9 = 0.0
tmp10 = tmp8 > tmp9
tmp11 = 0.01
tmp12 = tmp8 * tmp11
tmp13 = tl.where(tmp10, tmp8, tmp12)
tl.store(out_ptr1 + (y0 + 4 * x2 + 16 * y1), tmp13, xmask & ymask)
tl.store(out_ptr2 + (x2 + 4 * y3), tmp13, xmask & ymask)
@triton.jit
def triton_poi_fused_convolution_2(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl
.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = xindex // 4 % 4
tmp0 = tl.load(in_out_ptr0 + x3, xmask)
tmp1 = tl.load(in_ptr0 + x1, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + x3, tmp2, xmask)
@triton.jit
def triton_poi_fused_convolution_3(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl
.constexpr):
xnumel = 80
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = xindex // 5 % 4
tmp0 = tl.load(in_out_ptr0 + x3, xmask)
tmp1 = tl.load(in_ptr0 + x1, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + x3, tmp2, xmask)
@triton.jit
def triton_poi_fused_native_layer_norm_4(in_ptr0, out_ptr0, out_ptr1,
xnumel, XBLOCK: tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 4
x1 = xindex // 4
x2 = xindex
tmp0 = tl.load(in_ptr0 + (x0 + 20 * x1), xmask)
tmp1 = tl.load(in_ptr0 + (5 + x0 + 20 * x1), xmask)
tmp3 = tl.load(in_ptr0 + (10 + x0 + 20 * x1), xmask)
tmp5 = tl.load(in_ptr0 + (15 + x0 + 20 * x1), xmask)
tmp2 = tmp0 + tmp1
tmp4 = tmp2 + tmp3
tmp6 = tmp4 + tmp5
tmp7 = 4.0
tmp8 = tmp6 / tmp7
tmp9 = tmp0 - tmp8
tmp10 = tmp9 * tmp9
tmp11 = tmp1 - tmp8
tmp12 = tmp11 * tmp11
tmp13 = tmp10 + tmp12
tmp14 = tmp3 - tmp8
tmp15 = tmp14 * tmp14
tmp16 = tmp13 + tmp15
tmp17 = tmp5 - tmp8
tmp18 = tmp17 * tmp17
tmp19 = tmp16 + tmp18
tmp20 = tmp19 / tmp7
tmp21 = 1e-05
tmp22 = tmp20 + tmp21
tmp23 = libdevice.rsqrt(tmp22)
tl.store(out_ptr0 + x2, tmp8, xmask)
tl.store(out_ptr1 + x2, tmp23, xmask)
@triton.jit
def triton_poi_fused_convolution_leaky_relu_native_layer_norm_5(in_ptr0,
in_ptr1, in_ptr2, in_ptr3, in_ptr4, out_ptr1, out_ptr2, out_ptr3,
ynumel, xnumel, YBLOCK: tl.constexpr, XBLOCK: tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y3 = yindex
y1 = yindex // 4
y0 = yindex % 4
tmp0 = tl.load(in_ptr0 + (x2 + 5 * y3), xmask & ymask, eviction_policy=
'evict_last')
tmp1 = tl.load(in_ptr1 + (x2 + 4 * y1), xmask & ymask, eviction_policy=
'evict_last')
tmp3 = tl.load(in_ptr2 + (x2 + 4 * y1), xmask & ymask, eviction_policy=
'evict_last')
tmp5 = tl.load(in_ptr3 + y0, ymask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr4 + y0, ymask, eviction_policy='evict_last')
tmp2 = tmp0 - tmp1
tmp4 = tmp2 * tmp3
tmp6 = tmp4 * tmp5
tmp8 = tmp6 + tmp7
tmp9 = 0.0
tmp10 = tmp8 > tmp9
tmp11 = 0.01
tmp12 = tmp8 * tmp11
tmp13 = tl.where(tmp10, tmp8, tmp12)
tl.store(out_ptr1 + (y0 + 4 * x2 + 16 * y1), tmp13, xmask & ymask)
tl.store(out_ptr2 + (x2 + 4 * y3), tmp13, xmask & ymask)
tl.store(out_ptr3 + (x2 + 4 * y3), tmp13, xmask & ymask)
@triton.jit
def triton_poi_fused_add_convolution_mul_sigmoid_6(in_out_ptr0, in_out_ptr1,
in_ptr0, in_ptr1, in_ptr2, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = xindex // 4 % 4
tmp0 = tl.load(in_out_ptr0 + x3, xmask)
tmp1 = tl.load(in_ptr0 + x1, xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_out_ptr1 + x3, xmask)
tmp4 = tl.load(in_ptr1 + x1, xmask, eviction_policy='evict_last')
tmp8 = tl.load(in_ptr2 + x3, xmask)
tmp2 = tmp0 + tmp1
tmp5 = tmp3 + tmp4
tmp6 = tl.sigmoid(tmp5)
tmp7 = tmp2 * tmp6
tmp9 = tmp7 + tmp8
tl.store(in_out_ptr0 + x3, tmp2, xmask)
tl.store(in_out_ptr1 + x3, tmp5, xmask)
tl.store(out_ptr0 + x3, tmp9, xmask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7, primals_8, primals_9, primals_10, primals_11, primals_12,
primals_13, primals_14, primals_15) = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_2, (4,), (1,))
assert_size_stride(primals_3, (4,), (1,))
assert_size_stride(primals_4, (4, 4, 1), (4, 1, 1))
assert_size_stride(primals_5, (4,), (1,))
assert_size_stride(primals_6, (4,), (1,))
assert_size_stride(primals_7, (4,), (1,))
assert_size_stride(primals_8, (4, 4, 2), (8, 2, 1))
assert_size_stride(primals_9, (4,), (1,))
assert_size_stride(primals_10, (4,), (1,))
assert_size_stride(primals_11, (4,), (1,))
assert_size_stride(primals_12, (4, 4, 1), (4, 1, 1))
assert_size_stride(primals_13, (4,), (1,))
assert_size_stride(primals_14, (4, 4, 1), (4, 1, 1))
assert_size_stride(primals_15, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 1), (4, 1, 16), torch.float32)
buf1 = empty_strided_cuda((4, 4, 1), (4, 1, 16), torch.float32)
get_raw_stream(0)
triton_poi_fused_native_layer_norm_0[grid(16)](primals_1, buf0,
buf1, 16, XBLOCK=16, num_warps=1, num_stages=1)
buf3 = empty_strided_cuda((4, 4, 4), (16, 1, 4), torch.float32)
buf4 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
triton_poi_fused_convolution_leaky_relu_native_layer_norm_1[grid(16, 4)
](primals_1, buf0, buf1, primals_2, primals_3, buf3, buf4, 16,
4, XBLOCK=4, YBLOCK=8, num_warps=1, num_stages=1)
buf5 = extern_kernels.convolution(buf4, primals_4, stride=(1,),
padding=(0,), dilation=(1,), transposed=False, output_padding=(
0,), groups=1, bias=None)
assert_size_stride(buf5, (4, 4, 4), (16, 4, 1))
buf6 = buf5
del buf5
triton_poi_fused_convolution_2[grid(64)](buf6, primals_5, 64,
XBLOCK=64, num_warps=1, num_stages=1)
del primals_5
buf7 = buf1
del buf1
buf8 = buf0
del buf0
triton_poi_fused_native_layer_norm_0[grid(16)](buf6, buf7, buf8, 16,
XBLOCK=16, num_warps=1, num_stages=1)
buf10 = reinterpret_tensor(buf4, (4, 4, 4), (16, 1, 4), 0)
del buf4
buf11 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
triton_poi_fused_convolution_leaky_relu_native_layer_norm_1[grid(16, 4)
](buf6, buf7, buf8, primals_6, primals_7, buf10, buf11, 16, 4,
XBLOCK=4, YBLOCK=8, num_warps=1, num_stages=1)
buf12 = extern_kernels.convolution(buf11, primals_8, stride=(1,),
padding=(1,), dilation=(1,), transposed=False, output_padding=(
0,), groups=1, bias=None)
assert_size_stride(buf12, (4, 4, 5), (20, 5, 1))
buf13 = buf12
del buf12
triton_poi_fused_convolution_3[grid(80)](buf13, primals_9, 80,
XBLOCK=128, num_warps=4, num_stages=1)
del primals_9
buf14 = buf8
del buf8
buf15 = buf7
del buf7
triton_poi_fused_native_layer_norm_4[grid(16)](buf13, buf14, buf15,
16, XBLOCK=16, num_warps=1, num_stages=1)
buf17 = reinterpret_tensor(buf11, (4, 4, 4), (16, 1, 4), 0)
del buf11
buf18 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
buf21 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
triton_poi_fused_convolution_leaky_relu_native_layer_norm_5[grid(16, 4)
](buf13, buf14, buf15, primals_10, primals_11, buf17, buf18,
buf21, 16, 4, XBLOCK=4, YBLOCK=16, num_warps=1, num_stages=1)
del buf14
del buf15
buf19 = extern_kernels.convolution(buf18, primals_12, stride=(1,),
padding=(0,), dilation=(1,), transposed=False, output_padding=(
0,), groups=1, bias=None)
assert_size_stride(buf19, (4, 4, 4), (16, 4, 1))
del buf18
buf22 = extern_kernels.convolution(buf21, primals_14, stride=(1,),
padding=(0,), dilation=(1,), transposed=False, output_padding=(
0,), groups=1, bias=None)
assert_size_stride(buf22, (4, 4, 4), (16, 4, 1))
buf20 = buf19
del buf19
buf23 = buf22
del buf22
buf24 = buf21
del buf21
triton_poi_fused_add_convolution_mul_sigmoid_6[grid(64)](buf20,
buf23, primals_13, primals_15, primals_1, buf24, 64, XBLOCK=64,
num_warps=1, num_stages=1)
del primals_13
del primals_15
return (buf24, primals_1, primals_2, primals_3, primals_4, primals_6,
primals_7, primals_8, primals_10, primals_11, primals_12,
primals_14, buf3, buf6, buf10, buf13, buf17, buf20, buf23)
class DilConv1dWithGLUNew(nn.Module):
def __init__(self, num_channels, dilation, lenght=100, kernel_size=2,
activation=F.leaky_relu, residual_connection=True, dropout=0.2):
super(DilConv1dWithGLUNew, self).__init__()
self.dilation = dilation
self.start_ln = nn.LayerNorm(num_channels)
self.start_conv1x1 = nn.Conv1d(num_channels, num_channels,
kernel_size=1)
self.dilconv_ln = nn.LayerNorm(num_channels)
self.dilated_conv = nn.Conv1d(num_channels, num_channels, dilation=
dilation, kernel_size=kernel_size, padding=dilation)
self.gate_ln = nn.LayerNorm(num_channels)
self.end_conv1x1 = nn.Conv1d(num_channels, num_channels, kernel_size=1)
self.gated_conv1x1 = nn.Conv1d(num_channels, num_channels,
kernel_size=1)
self.activation = activation
self.buffer = None
self.residual_connection = residual_connection
def clear_buffer(self):
self.buffer = None
def forward(self, input_0):
primals_2 = self.start_ln.weight
primals_3 = self.start_ln.bias
primals_4 = self.start_conv1x1.weight
primals_5 = self.start_conv1x1.bias
primals_6 = self.dilconv_ln.weight
primals_7 = self.dilconv_ln.bias
primals_8 = self.dilated_conv.weight
primals_9 = self.dilated_conv.bias
primals_10 = self.gate_ln.weight
primals_11 = self.gate_ln.bias
primals_12 = self.end_conv1x1.weight
primals_13 = self.end_conv1x1.bias
primals_14 = self.gated_conv1x1.weight
primals_15 = self.gated_conv1x1.bias
primals_1 = input_0
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7, primals_8, primals_9,
primals_10, primals_11, primals_12, primals_13, primals_14,
primals_15])
return output[0]
| icyray/proGENTRL | DilConv1dWithGLU | false | 6,880 | [
"MIT"
] | 1 | c48305c3411ecb604c4f26f5e6b62f285e42e696 | https://github.com/icyray/proGENTRL/tree/c48305c3411ecb604c4f26f5e6b62f285e42e696 | import torch
import torch.nn as nn
import torch.nn.functional as F
class Model(nn.Module):
def __init__(self, num_channels, dilation, lenght=100, kernel_size=2,
activation=F.leaky_relu, residual_connection=True, dropout=0.2):
super().__init__()
self.dilation = dilation
self.start_ln = nn.LayerNorm(num_channels)
self.start_conv1x1 = nn.Conv1d(num_channels, num_channels,
kernel_size=1)
self.dilconv_ln = nn.LayerNorm(num_channels)
self.dilated_conv = nn.Conv1d(num_channels, num_channels, dilation=
dilation, kernel_size=kernel_size, padding=dilation)
self.gate_ln = nn.LayerNorm(num_channels)
self.end_conv1x1 = nn.Conv1d(num_channels, num_channels, kernel_size=1)
self.gated_conv1x1 = nn.Conv1d(num_channels, num_channels,
kernel_size=1)
self.activation = activation
self.buffer = None
self.residual_connection = residual_connection
def clear_buffer(self):
self.buffer = None
def forward(self, x_inp, sampling=False):
x = self.start_ln(x_inp.transpose(1, 2)).transpose(1, 2)
x = self.activation(x)
x = self.start_conv1x1(x)
x = self.dilconv_ln(x.transpose(1, 2)).transpose(1, 2)
x = self.activation(x)
if sampling:
if self.buffer is None:
self.buffer = x
else:
pre_buffer = torch.cat([self.buffer, x], dim=2)
self.buffer = pre_buffer[:, :, -(self.dilation + 1):]
if self.buffer.shape[2] == self.dilation + 1:
x = self.buffer
else:
x = torch.cat([torch.zeros(self.buffer.shape[0], self.
buffer.shape[1], self.dilation + 1 - self.buffer.shape[
2], device=x_inp.device), self.buffer], dim=2)
x = self.dilated_conv(x)[:, :, self.dilation:]
x = x[:, :, :x_inp.shape[-1]]
else:
x = self.dilated_conv(x)[:, :, :x_inp.shape[-1]]
x = self.gate_ln(x.transpose(1, 2)).transpose(1, 2)
x = self.activation(x)
x = self.end_conv1x1(x) * torch.sigmoid(self.gated_conv1x1(x))
if self.residual_connection:
x = x + x_inp
return x
def get_inputs():
return [torch.rand([4, 4, 4])]
def get_init_inputs():
return [4, 1]
|
CausalConv2d | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/2a/c2aucoxj5ek5nl3pl5n67fq4vdl55x2rw66tvv2aq23fsb3jvf5q.py
# Topologically Sorted Source Nodes: [out], Original ATen: [aten.constant_pad_nd]
# Source node to ATen node mapping:
# out => constant_pad_nd
# Graph fragment:
# %constant_pad_nd : [num_users=2] = call_function[target=torch.ops.aten.constant_pad_nd.default](args = (%primals_1, [3, 0, 3, 0], 0.0), kwargs = {})
triton_poi_fused_constant_pad_nd_0 = async_compile.triton('triton_poi_fused_constant_pad_nd_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[1024],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_constant_pad_nd_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_constant_pad_nd_0(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 784
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = (xindex // 7) % 7
x0 = xindex % 7
x2 = (xindex // 49)
x4 = xindex
tmp0 = (-3) + x1
tmp1 = tl.full([1], 0, tl.int64)
tmp2 = tmp0 >= tmp1
tmp3 = (-3) + x0
tmp4 = tmp3 >= tmp1
tmp5 = tmp2 & tmp4
tmp6 = tl.load(in_ptr0 + ((-15) + x0 + (4*x1) + (16*x2)), tmp5 & xmask, other=0.0)
tl.store(out_ptr0 + (x4), tmp6, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/6r/c6rjd3rxthhw2ub6d2gtgjazrjeoyhalmj36ujwzgwexsis27e73.py
# Topologically Sorted Source Nodes: [_weight_norm], Original ATen: [aten._weight_norm_interface]
# Source node to ATen node mapping:
# _weight_norm => div, mul, pow_1, pow_2, sum_1
# Graph fragment:
# %pow_1 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%primals_3, 2), kwargs = {})
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%pow_1, [1, 2, 3], True), kwargs = {})
# %pow_2 : [num_users=2] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%sum_1, 0.5), kwargs = {})
# %div : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%primals_2, %pow_2), kwargs = {})
# %mul : [num_users=2] = call_function[target=torch.ops.aten.mul.Tensor](args = (%primals_3, %div), kwargs = {})
triton_per_fused__weight_norm_interface_1 = async_compile.triton('triton_per_fused__weight_norm_interface_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[4, 64],
reduction_hint=ReductionHint.INNER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32', 5: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 5), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused__weight_norm_interface_1', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 1, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused__weight_norm_interface_1(in_out_ptr0, in_ptr0, in_ptr1, out_ptr0, xnumel, rnumel, XBLOCK : tl.constexpr):
xnumel = 4
rnumel = 64
RBLOCK: tl.constexpr = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
roffset = 0
rmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
r1 = rindex
x0 = xindex
tmp0 = tl.load(in_ptr0 + (r1 + (64*x0)), xmask, other=0.0)
tmp7 = tl.load(in_ptr1 + (x0), xmask, eviction_policy='evict_last')
tmp1 = tmp0 * tmp0
tmp2 = tl.broadcast_to(tmp1, [XBLOCK, RBLOCK])
tmp4 = tl.where(xmask, tmp2, 0)
tmp5 = tl.sum(tmp4, 1)[:, None]
tmp6 = libdevice.sqrt(tmp5)
tmp8 = tmp7 / tmp6
tmp9 = tmp0 * tmp8
tl.debug_barrier()
tl.store(in_out_ptr0 + (x0), tmp6, xmask)
tl.store(out_ptr0 + (r1 + (64*x0)), tmp9, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/w5/cw5gytijzzkwnfpq2a2axdsj4pfxgxmwiuzizuyd4bw5uwnanzw7.py
# Topologically Sorted Source Nodes: [out_1], Original ATen: [aten.convolution]
# Source node to ATen node mapping:
# out_1 => convolution
# Graph fragment:
# %convolution : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%constant_pad_nd, %mul, %primals_4, [1, 1], [0, 0], [1, 1], False, [0, 0], 1), kwargs = {})
triton_poi_fused_convolution_2 = async_compile.triton('triton_poi_fused_convolution_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_2', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_2(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = (xindex // 16) % 4
tmp0 = tl.load(in_out_ptr0 + (x3), xmask)
tmp1 = tl.load(in_ptr0 + (x1), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + (x3), tmp2, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (4, 1, 1, 1), (1, 1, 1, 1))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_4, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 7, 7), (196, 49, 7, 1), torch.float32)
# Topologically Sorted Source Nodes: [out], Original ATen: [aten.constant_pad_nd]
stream0 = get_raw_stream(0)
triton_poi_fused_constant_pad_nd_0.run(primals_1, buf0, 784, grid=grid(784), stream=stream0)
del primals_1
buf1 = empty_strided_cuda((4, 1, 1, 1), (1, 4, 4, 4), torch.float32)
buf2 = reinterpret_tensor(buf1, (4, 1, 1, 1), (1, 1, 1, 1), 0); del buf1 # reuse
buf3 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [_weight_norm], Original ATen: [aten._weight_norm_interface]
triton_per_fused__weight_norm_interface_1.run(buf2, primals_3, primals_2, buf3, 4, 64, grid=grid(4), stream=stream0)
# Topologically Sorted Source Nodes: [out_1], Original ATen: [aten.convolution]
buf4 = extern_kernels.convolution(buf0, buf3, stride=(1, 1), padding=(0, 0), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf4, (4, 4, 4, 4), (64, 16, 4, 1))
buf5 = buf4; del buf4 # reuse
# Topologically Sorted Source Nodes: [out_1], Original ATen: [aten.convolution]
triton_poi_fused_convolution_2.run(buf5, primals_4, 256, grid=grid(256), stream=stream0)
del primals_4
return (buf5, buf3, primals_2, primals_3, buf0, buf2, buf3, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 1, 1, 1), (1, 1, 1, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
class WNConv2d(nn.Module):
def __init__(self, in_channel, out_channel, kernel_size, stride=1,
padding=0, bias=True, activation=None):
super().__init__()
self.conv = nn.utils.weight_norm(nn.Conv2d(in_channel, out_channel,
kernel_size, stride=stride, padding=padding, bias=bias))
self.out_channel = out_channel
if isinstance(kernel_size, int):
kernel_size = [kernel_size, kernel_size]
self.kernel_size = kernel_size
self.activation = activation
def forward(self, input):
out = self.conv(input)
if self.activation is not None:
out = self.activation(out)
return out
class CausalConv2d(nn.Module):
def __init__(self, in_channel, out_channel, kernel_size, stride=1,
padding='downright', activation=None):
super().__init__()
if isinstance(kernel_size, int):
kernel_size = [kernel_size] * 2
self.kernel_size = kernel_size
if padding == 'downright':
pad = [kernel_size[1] - 1, 0, kernel_size[0] - 1, 0]
elif padding == 'down' or padding == 'causal':
pad = kernel_size[1] // 2
pad = [pad, pad, kernel_size[0] - 1, 0]
self.causal = 0
if padding == 'causal':
self.causal = kernel_size[1] // 2
self.pad = nn.ZeroPad2d(pad)
self.conv = WNConv2d(in_channel, out_channel, kernel_size, stride=
stride, padding=0, activation=activation)
def forward(self, input):
out = self.pad(input)
if self.causal > 0:
self.conv.conv.weight_v.data[:, :, -1, self.causal:].zero_()
out = self.conv(out)
return out
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'in_channel': 4, 'out_channel': 4, 'kernel_size': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime.triton_helpers import libdevice
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_constant_pad_nd_0(in_ptr0, out_ptr0, xnumel, XBLOCK:
tl.constexpr):
xnumel = 784
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = xindex // 7 % 7
x0 = xindex % 7
x2 = xindex // 49
x4 = xindex
tmp0 = -3 + x1
tmp1 = tl.full([1], 0, tl.int64)
tmp2 = tmp0 >= tmp1
tmp3 = -3 + x0
tmp4 = tmp3 >= tmp1
tmp5 = tmp2 & tmp4
tmp6 = tl.load(in_ptr0 + (-15 + x0 + 4 * x1 + 16 * x2), tmp5 & xmask,
other=0.0)
tl.store(out_ptr0 + x4, tmp6, xmask)
@triton.jit
def triton_per_fused__weight_norm_interface_1(in_out_ptr0, in_ptr0, in_ptr1,
out_ptr0, xnumel, rnumel, XBLOCK: tl.constexpr):
xnumel = 4
RBLOCK: tl.constexpr = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
r1 = rindex
x0 = xindex
tmp0 = tl.load(in_ptr0 + (r1 + 64 * x0), xmask, other=0.0)
tmp7 = tl.load(in_ptr1 + x0, xmask, eviction_policy='evict_last')
tmp1 = tmp0 * tmp0
tmp2 = tl.broadcast_to(tmp1, [XBLOCK, RBLOCK])
tmp4 = tl.where(xmask, tmp2, 0)
tmp5 = tl.sum(tmp4, 1)[:, None]
tmp6 = libdevice.sqrt(tmp5)
tmp8 = tmp7 / tmp6
tmp9 = tmp0 * tmp8
tl.debug_barrier()
tl.store(in_out_ptr0 + x0, tmp6, xmask)
tl.store(out_ptr0 + (r1 + 64 * x0), tmp9, xmask)
@triton.jit
def triton_poi_fused_convolution_2(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl
.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = xindex // 16 % 4
tmp0 = tl.load(in_out_ptr0 + x3, xmask)
tmp1 = tl.load(in_ptr0 + x1, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + x3, tmp2, xmask)
def call(args):
primals_1, primals_2, primals_3, primals_4 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (4, 1, 1, 1), (1, 1, 1, 1))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_4, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 7, 7), (196, 49, 7, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_constant_pad_nd_0[grid(784)](primals_1, buf0, 784,
XBLOCK=256, num_warps=4, num_stages=1)
del primals_1
buf1 = empty_strided_cuda((4, 1, 1, 1), (1, 4, 4, 4), torch.float32)
buf2 = reinterpret_tensor(buf1, (4, 1, 1, 1), (1, 1, 1, 1), 0)
del buf1
buf3 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
triton_per_fused__weight_norm_interface_1[grid(4)](buf2, primals_3,
primals_2, buf3, 4, 64, XBLOCK=1, num_warps=2, num_stages=1)
buf4 = extern_kernels.convolution(buf0, buf3, stride=(1, 1),
padding=(0, 0), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf4, (4, 4, 4, 4), (64, 16, 4, 1))
buf5 = buf4
del buf4
triton_poi_fused_convolution_2[grid(256)](buf5, primals_4, 256,
XBLOCK=256, num_warps=4, num_stages=1)
del primals_4
return buf5, buf3, primals_2, primals_3, buf0, buf2, buf3
class WNConv2d(nn.Module):
def __init__(self, in_channel, out_channel, kernel_size, stride=1,
padding=0, bias=True, activation=None):
super().__init__()
self.conv = nn.utils.weight_norm(nn.Conv2d(in_channel, out_channel,
kernel_size, stride=stride, padding=padding, bias=bias))
self.out_channel = out_channel
if isinstance(kernel_size, int):
kernel_size = [kernel_size, kernel_size]
self.kernel_size = kernel_size
self.activation = activation
def forward(self, input):
out = self.conv(input)
if self.activation is not None:
out = self.activation(out)
return out
class CausalConv2dNew(nn.Module):
def __init__(self, in_channel, out_channel, kernel_size, stride=1,
padding='downright', activation=None):
super().__init__()
if isinstance(kernel_size, int):
kernel_size = [kernel_size] * 2
self.kernel_size = kernel_size
if padding == 'downright':
pad = [kernel_size[1] - 1, 0, kernel_size[0] - 1, 0]
elif padding == 'down' or padding == 'causal':
pad = kernel_size[1] // 2
pad = [pad, pad, kernel_size[0] - 1, 0]
self.causal = 0
if padding == 'causal':
self.causal = kernel_size[1] // 2
self.pad = nn.ZeroPad2d(pad)
self.conv = WNConv2d(in_channel, out_channel, kernel_size, stride=
stride, padding=0, activation=activation)
def forward(self, input_0):
primals_4 = self.conv.conv.bias
primals_2 = self.conv.conv.weight_g
primals_1 = self.conv.conv.weight_v
primals_3 = input_0
output = call([primals_1, primals_2, primals_3, primals_4])
return output[0]
| imatge-upc/pixelcoordEDL | CausalConv2d | false | 6,881 | [
"MIT"
] | 1 | 353632feed6ac8c93758c1a2a1b7a477e7ff053c | https://github.com/imatge-upc/pixelcoordEDL/tree/353632feed6ac8c93758c1a2a1b7a477e7ff053c | import torch
import torch.nn as nn
class WNConv2d(nn.Module):
def __init__(self, in_channel, out_channel, kernel_size, stride=1,
padding=0, bias=True, activation=None):
super().__init__()
self.conv = nn.utils.weight_norm(nn.Conv2d(in_channel, out_channel,
kernel_size, stride=stride, padding=padding, bias=bias))
self.out_channel = out_channel
if isinstance(kernel_size, int):
kernel_size = [kernel_size, kernel_size]
self.kernel_size = kernel_size
self.activation = activation
def forward(self, input):
out = self.conv(input)
if self.activation is not None:
out = self.activation(out)
return out
class Model(nn.Module):
def __init__(self, in_channel, out_channel, kernel_size, stride=1,
padding='downright', activation=None):
super().__init__()
if isinstance(kernel_size, int):
kernel_size = [kernel_size] * 2
self.kernel_size = kernel_size
if padding == 'downright':
pad = [kernel_size[1] - 1, 0, kernel_size[0] - 1, 0]
elif padding == 'down' or padding == 'causal':
pad = kernel_size[1] // 2
pad = [pad, pad, kernel_size[0] - 1, 0]
self.causal = 0
if padding == 'causal':
self.causal = kernel_size[1] // 2
self.pad = nn.ZeroPad2d(pad)
self.conv = WNConv2d(in_channel, out_channel, kernel_size, stride=
stride, padding=0, activation=activation)
def forward(self, input):
out = self.pad(input)
if self.causal > 0:
self.conv.conv.weight_v.data[:, :, -1, self.causal:].zero_()
out = self.conv(out)
return out
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4, 4, 4]
|
DY_Conv2d | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/l3/cl35tzbhrd24dhunkbb6gjs54aklpyr46oikqhoylcgmkcmhujil.py
# Topologically Sorted Source Nodes: [atten], Original ATen: [aten.mean]
# Source node to ATen node mapping:
# atten => mean
# Graph fragment:
# %mean : [num_users=2] = call_function[target=torch.ops.aten.mean.dim](args = (%primals_1, [2, 3], True), kwargs = {})
triton_per_fused_mean_0 = async_compile.triton('triton_per_fused_mean_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[16, 16],
reduction_hint=ReductionHint.INNER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused_mean_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 1, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused_mean_0(in_out_ptr0, in_ptr0, xnumel, rnumel, XBLOCK : tl.constexpr):
xnumel = 16
rnumel = 16
RBLOCK: tl.constexpr = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
roffset = 0
rmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
r1 = rindex
x0 = xindex
tmp0 = tl.load(in_ptr0 + (r1 + (16*x0)), xmask, other=0.0)
tmp1 = tl.broadcast_to(tmp0, [XBLOCK, RBLOCK])
tmp3 = tl.where(xmask, tmp1, 0)
tmp4 = tl.sum(tmp3, 1)[:, None]
tmp5 = 16.0
tmp6 = tmp4 / tmp5
tl.debug_barrier()
tl.store(in_out_ptr0 + (x0), tmp6, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/ad/cadccuyhl7stcp3nyqfgohiwbiv5ckfzxsye27ithwsill6dvmh4.py
# Topologically Sorted Source Nodes: [atten_1, atten_2], Original ATen: [aten.convolution, aten.relu]
# Source node to ATen node mapping:
# atten_1 => convolution
# atten_2 => relu
# Graph fragment:
# %convolution : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%mean, %primals_2, %primals_3, [1, 1], [0, 0], [1, 1], False, [0, 0], 1), kwargs = {})
# %relu : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%convolution,), kwargs = {})
triton_poi_fused_convolution_relu_1 = async_compile.triton('triton_poi_fused_convolution_relu_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[4],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_relu_1', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_relu_1(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 4
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_out_ptr0 + (x0), xmask)
tmp1 = tl.load(in_ptr0 + (0))
tmp2 = tl.broadcast_to(tmp1, [XBLOCK])
tmp3 = tmp0 + tmp2
tmp4 = tl.full([1], 0, tl.int32)
tmp5 = triton_helpers.maximum(tmp4, tmp3)
tl.store(in_out_ptr0 + (x0), tmp5, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/k2/ck2mamkqpmuzem4n3p4ij6fmfpy2bcbblg6sx6wwslgqwuqq5ifh.py
# Topologically Sorted Source Nodes: [atten_3], Original ATen: [aten.convolution]
# Source node to ATen node mapping:
# atten_3 => convolution_1
# Graph fragment:
# %convolution_1 : [num_users=3] = call_function[target=torch.ops.aten.convolution.default](args = (%relu, %primals_4, %primals_5, [1, 1], [0, 0], [1, 1], False, [0, 0], 1), kwargs = {})
triton_poi_fused_convolution_2 = async_compile.triton('triton_poi_fused_convolution_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_2', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_2(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + (x2), tmp2, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/lt/cltwbpokq7b7gvah2tjf27qlzw6vpmwfuzs3xfk7mhbxym753kvi.py
# Topologically Sorted Source Nodes: [softmax], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# softmax => amax, exp, sub
# Graph fragment:
# %amax : [num_users=1] = call_function[target=torch.ops.aten.amax.default](args = (%convolution_1, [1], True), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%convolution_1, %amax), kwargs = {})
# %exp : [num_users=2] = call_function[target=torch.ops.aten.exp.default](args = (%sub,), kwargs = {})
triton_poi_fused__softmax_3 = async_compile.triton('triton_poi_fused__softmax_3', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_3', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_3(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp3 = triton_helpers.maximum(tmp1, tmp2)
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp7 = triton_helpers.maximum(tmp5, tmp6)
tmp8 = tmp0 - tmp7
tmp9 = tl_math.exp(tmp8)
tl.store(out_ptr0 + (x2), tmp9, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/rr/crrmj7r54x5uk325xkhuskxp4m5prz3fpx53yc2st4o5pwbhq32p.py
# Topologically Sorted Source Nodes: [softmax], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# softmax => div, sum_1
# Graph fragment:
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%exp, [1], True), kwargs = {})
# %div : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%exp, %sum_1), kwargs = {})
triton_poi_fused__softmax_4 = async_compile.triton('triton_poi_fused__softmax_4', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_4', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_4(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + (x2), tmp8, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/to/ctogavafceeroff63wmlh3nv62ggyc3xrvpxvw55s6umuxptsi5q.py
# Topologically Sorted Source Nodes: [mul, W_1], Original ATen: [aten.mul, aten.sum]
# Source node to ATen node mapping:
# W_1 => sum_2
# mul => mul
# Graph fragment:
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%view_1, %view_2), kwargs = {})
# %sum_2 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%mul, [1]), kwargs = {})
triton_poi_fused_mul_sum_5 = async_compile.triton('triton_poi_fused_mul_sum_5', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[1024],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_mul_sum_5', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 8, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_mul_sum_5(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 576
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 144
x1 = (xindex // 144)
x2 = xindex
tmp0 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (4*x1), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr0 + (144 + x0), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr1 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr0 + (288 + x0), xmask, eviction_policy='evict_last')
tmp8 = tl.load(in_ptr1 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp11 = tl.load(in_ptr0 + (432 + x0), xmask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr1 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp2 = tmp0 * tmp1
tmp5 = tmp3 * tmp4
tmp6 = tmp2 + tmp5
tmp9 = tmp7 * tmp8
tmp10 = tmp6 + tmp9
tmp13 = tmp11 * tmp12
tmp14 = tmp10 + tmp13
tl.store(out_ptr0 + (x2), tmp14, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (1, 4, 1, 1), (4, 1, 1, 1))
assert_size_stride(primals_3, (1, ), (1, ))
assert_size_stride(primals_4, (4, 1, 1, 1), (1, 1, 1, 1))
assert_size_stride(primals_5, (4, ), (1, ))
assert_size_stride(primals_6, (16, 4, 3, 3), (36, 9, 3, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 1, 1), (4, 1, 16, 16), torch.float32)
buf1 = reinterpret_tensor(buf0, (4, 4, 1, 1), (4, 1, 1, 1), 0); del buf0 # reuse
# Topologically Sorted Source Nodes: [atten], Original ATen: [aten.mean]
stream0 = get_raw_stream(0)
triton_per_fused_mean_0.run(buf1, primals_1, 16, 16, grid=grid(16), stream=stream0)
# Topologically Sorted Source Nodes: [atten_1], Original ATen: [aten.convolution]
buf2 = extern_kernels.convolution(buf1, primals_2, stride=(1, 1), padding=(0, 0), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf2, (4, 1, 1, 1), (1, 1, 1, 1))
buf3 = buf2; del buf2 # reuse
# Topologically Sorted Source Nodes: [atten_1, atten_2], Original ATen: [aten.convolution, aten.relu]
triton_poi_fused_convolution_relu_1.run(buf3, primals_3, 4, grid=grid(4), stream=stream0)
del primals_3
# Topologically Sorted Source Nodes: [atten_3], Original ATen: [aten.convolution]
buf4 = extern_kernels.convolution(buf3, primals_4, stride=(1, 1), padding=(0, 0), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf4, (4, 4, 1, 1), (4, 1, 1, 1))
buf5 = buf4; del buf4 # reuse
# Topologically Sorted Source Nodes: [atten_3], Original ATen: [aten.convolution]
triton_poi_fused_convolution_2.run(buf5, primals_5, 16, grid=grid(16), stream=stream0)
del primals_5
buf6 = empty_strided_cuda((4, 4, 1, 1), (4, 1, 16, 16), torch.float32)
# Topologically Sorted Source Nodes: [softmax], Original ATen: [aten._softmax]
triton_poi_fused__softmax_3.run(buf5, buf6, 16, grid=grid(16), stream=stream0)
buf7 = empty_strided_cuda((4, 4, 1, 1), (4, 1, 16, 16), torch.float32)
# Topologically Sorted Source Nodes: [softmax], Original ATen: [aten._softmax]
triton_poi_fused__softmax_4.run(buf6, buf7, 16, grid=grid(16), stream=stream0)
del buf6
buf8 = empty_strided_cuda((4, 4, 4, 3, 3), (144, 36, 9, 3, 1), torch.float32)
# Topologically Sorted Source Nodes: [mul, W_1], Original ATen: [aten.mul, aten.sum]
triton_poi_fused_mul_sum_5.run(primals_6, buf7, buf8, 576, grid=grid(576), stream=stream0)
del buf7
# Topologically Sorted Source Nodes: [out], Original ATen: [aten.convolution]
buf9 = extern_kernels.convolution(reinterpret_tensor(primals_1, (1, 16, 4, 4), (256, 16, 4, 1), 0), reinterpret_tensor(buf8, (16, 4, 3, 3), (36, 9, 3, 1), 0), stride=(1, 1), padding=(1, 1), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=4, bias=None)
assert_size_stride(buf9, (1, 16, 4, 4), (256, 16, 4, 1))
return (reinterpret_tensor(buf9, (4, 4, 4, 4), (64, 16, 4, 1), 0), primals_2, primals_4, primals_6, buf1, buf3, buf5, reinterpret_tensor(buf8, (16, 4, 3, 3), (36, 9, 3, 1), 0), reinterpret_tensor(primals_1, (1, 16, 4, 4), (256, 16, 4, 1), 0), )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((1, 4, 1, 1), (4, 1, 1, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((1, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, 1, 1, 1), (1, 1, 1, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((16, 4, 3, 3), (36, 9, 3, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.nn.functional as F
class DY_Conv2d(nn.Conv2d):
def __init__(self, in_chan, out_chan, kernel_size=3, stride=1, padding=
1, dilation=1, groups=1, bias=False, act=nn.ReLU(inplace=True), K=4,
temperature=30, temp_anneal_steps=3000):
super(DY_Conv2d, self).__init__(in_chan, out_chan * K, kernel_size,
stride=stride, padding=padding, dilation=dilation, groups=
groups, bias=bias)
assert in_chan // 4 > 0
self.K = K
self.act = act
self.se_conv1 = nn.Conv2d(in_chan, in_chan // 4, 1, 1, 0, bias=True)
self.se_conv2 = nn.Conv2d(in_chan // 4, K, 1, 1, 0, bias=True)
self.temperature = temperature
self.temp_anneal_steps = temp_anneal_steps
self.temp_interval = (temperature - 1) / temp_anneal_steps
def get_atten(self, x):
bs, _, _h, _w = x.size()
atten = torch.mean(x, dim=(2, 3), keepdim=True)
atten = self.se_conv1(atten)
atten = self.act(atten)
atten = self.se_conv2(atten)
if self.training and self.temp_anneal_steps > 0:
atten = atten / self.temperature
self.temperature -= self.temp_interval
self.temp_anneal_steps -= 1
atten = atten.softmax(dim=1).view(bs, -1)
return atten
def forward(self, x):
bs, _, h, w = x.size()
atten = self.get_atten(x)
_out_chan, in_chan, k1, k2 = self.weight.size()
W = self.weight.view(1, self.K, -1, in_chan, k1, k2)
W = (W * atten.view(bs, self.K, 1, 1, 1, 1)).sum(dim=1)
W = W.view(-1, in_chan, k1, k2)
b = self.bias
if b is not None:
b = b.view(1, self.K, -1)
b = (b * atten.view(bs, self.K, 1)).sum(dim=1).view(-1)
x = x.view(1, -1, h, w)
out = F.conv2d(x, W, b, self.stride, self.padding, self.dilation,
self.groups * bs)
out = out.view(bs, -1, out.size(2), out.size(3))
return out
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'in_chan': 4, 'out_chan': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import math as tl_math
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_per_fused_mean_0(in_out_ptr0, in_ptr0, xnumel, rnumel, XBLOCK:
tl.constexpr):
xnumel = 16
RBLOCK: tl.constexpr = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
r1 = rindex
x0 = xindex
tmp0 = tl.load(in_ptr0 + (r1 + 16 * x0), xmask, other=0.0)
tmp1 = tl.broadcast_to(tmp0, [XBLOCK, RBLOCK])
tmp3 = tl.where(xmask, tmp1, 0)
tmp4 = tl.sum(tmp3, 1)[:, None]
tmp5 = 16.0
tmp6 = tmp4 / tmp5
tl.debug_barrier()
tl.store(in_out_ptr0 + x0, tmp6, xmask)
@triton.jit
def triton_poi_fused_convolution_relu_1(in_out_ptr0, in_ptr0, xnumel,
XBLOCK: tl.constexpr):
xnumel = 4
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_out_ptr0 + x0, xmask)
tmp1 = tl.load(in_ptr0 + 0)
tmp2 = tl.broadcast_to(tmp1, [XBLOCK])
tmp3 = tmp0 + tmp2
tmp4 = tl.full([1], 0, tl.int32)
tmp5 = triton_helpers.maximum(tmp4, tmp3)
tl.store(in_out_ptr0 + x0, tmp5, xmask)
@triton.jit
def triton_poi_fused_convolution_2(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl
.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + x2, tmp2, xmask)
@triton.jit
def triton_poi_fused__softmax_3(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp3 = triton_helpers.maximum(tmp1, tmp2)
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp7 = triton_helpers.maximum(tmp5, tmp6)
tmp8 = tmp0 - tmp7
tmp9 = tl_math.exp(tmp8)
tl.store(out_ptr0 + x2, tmp9, xmask)
@triton.jit
def triton_poi_fused__softmax_4(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + x2, tmp8, xmask)
@triton.jit
def triton_poi_fused_mul_sum_5(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK:
tl.constexpr):
xnumel = 576
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 144
x1 = xindex // 144
x2 = xindex
tmp0 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + 4 * x1, xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr0 + (144 + x0), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr1 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr0 + (288 + x0), xmask, eviction_policy='evict_last')
tmp8 = tl.load(in_ptr1 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp11 = tl.load(in_ptr0 + (432 + x0), xmask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr1 + (3 + 4 * x1), xmask, eviction_policy='evict_last'
)
tmp2 = tmp0 * tmp1
tmp5 = tmp3 * tmp4
tmp6 = tmp2 + tmp5
tmp9 = tmp7 * tmp8
tmp10 = tmp6 + tmp9
tmp13 = tmp11 * tmp12
tmp14 = tmp10 + tmp13
tl.store(out_ptr0 + x2, tmp14, xmask)
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (1, 4, 1, 1), (4, 1, 1, 1))
assert_size_stride(primals_3, (1,), (1,))
assert_size_stride(primals_4, (4, 1, 1, 1), (1, 1, 1, 1))
assert_size_stride(primals_5, (4,), (1,))
assert_size_stride(primals_6, (16, 4, 3, 3), (36, 9, 3, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 1, 1), (4, 1, 16, 16), torch.float32)
buf1 = reinterpret_tensor(buf0, (4, 4, 1, 1), (4, 1, 1, 1), 0)
del buf0
get_raw_stream(0)
triton_per_fused_mean_0[grid(16)](buf1, primals_1, 16, 16, XBLOCK=8,
num_warps=2, num_stages=1)
buf2 = extern_kernels.convolution(buf1, primals_2, stride=(1, 1),
padding=(0, 0), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf2, (4, 1, 1, 1), (1, 1, 1, 1))
buf3 = buf2
del buf2
triton_poi_fused_convolution_relu_1[grid(4)](buf3, primals_3, 4,
XBLOCK=4, num_warps=1, num_stages=1)
del primals_3
buf4 = extern_kernels.convolution(buf3, primals_4, stride=(1, 1),
padding=(0, 0), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf4, (4, 4, 1, 1), (4, 1, 1, 1))
buf5 = buf4
del buf4
triton_poi_fused_convolution_2[grid(16)](buf5, primals_5, 16,
XBLOCK=16, num_warps=1, num_stages=1)
del primals_5
buf6 = empty_strided_cuda((4, 4, 1, 1), (4, 1, 16, 16), torch.float32)
triton_poi_fused__softmax_3[grid(16)](buf5, buf6, 16, XBLOCK=16,
num_warps=1, num_stages=1)
buf7 = empty_strided_cuda((4, 4, 1, 1), (4, 1, 16, 16), torch.float32)
triton_poi_fused__softmax_4[grid(16)](buf6, buf7, 16, XBLOCK=16,
num_warps=1, num_stages=1)
del buf6
buf8 = empty_strided_cuda((4, 4, 4, 3, 3), (144, 36, 9, 3, 1),
torch.float32)
triton_poi_fused_mul_sum_5[grid(576)](primals_6, buf7, buf8, 576,
XBLOCK=128, num_warps=4, num_stages=1)
del buf7
buf9 = extern_kernels.convolution(reinterpret_tensor(primals_1, (1,
16, 4, 4), (256, 16, 4, 1), 0), reinterpret_tensor(buf8, (16, 4,
3, 3), (36, 9, 3, 1), 0), stride=(1, 1), padding=(1, 1),
dilation=(1, 1), transposed=False, output_padding=(0, 0),
groups=4, bias=None)
assert_size_stride(buf9, (1, 16, 4, 4), (256, 16, 4, 1))
return reinterpret_tensor(buf9, (4, 4, 4, 4), (64, 16, 4, 1), 0
), primals_2, primals_4, primals_6, buf1, buf3, buf5, reinterpret_tensor(
buf8, (16, 4, 3, 3), (36, 9, 3, 1), 0), reinterpret_tensor(primals_1,
(1, 16, 4, 4), (256, 16, 4, 1), 0)
class DY_Conv2dNew(nn.Conv2d):
def __init__(self, in_chan, out_chan, kernel_size=3, stride=1, padding=
1, dilation=1, groups=1, bias=False, act=nn.ReLU(inplace=True), K=4,
temperature=30, temp_anneal_steps=3000):
super(DY_Conv2dNew, self).__init__(in_chan, out_chan * K,
kernel_size, stride=stride, padding=padding, dilation=dilation,
groups=groups, bias=bias)
assert in_chan // 4 > 0
self.K = K
self.act = act
self.se_conv1 = nn.Conv2d(in_chan, in_chan // 4, 1, 1, 0, bias=True)
self.se_conv2 = nn.Conv2d(in_chan // 4, K, 1, 1, 0, bias=True)
self.temperature = temperature
self.temp_anneal_steps = temp_anneal_steps
self.temp_interval = (temperature - 1) / temp_anneal_steps
def get_atten(self, x):
bs, _, _h, _w = x.size()
atten = torch.mean(x, dim=(2, 3), keepdim=True)
atten = self.se_conv1(atten)
atten = self.act(atten)
atten = self.se_conv2(atten)
if self.training and self.temp_anneal_steps > 0:
atten = atten / self.temperature
self.temperature -= self.temp_interval
self.temp_anneal_steps -= 1
atten = atten.softmax(dim=1).view(bs, -1)
return atten
def forward(self, input_0):
primals_6 = self.weight
primals_2 = self.se_conv1.weight
primals_3 = self.se_conv1.bias
primals_4 = self.se_conv2.weight
primals_5 = self.se_conv2.bias
primals_1 = input_0
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6])
return output[0]
| imvladikon/pytorch-loss | DY_Conv2d | false | 6,882 | [
"MIT"
] | 1 | 6cfaabe1be898e1ff000b3dffb46d0ef09096f6b | https://github.com/imvladikon/pytorch-loss/tree/6cfaabe1be898e1ff000b3dffb46d0ef09096f6b | import torch
import torch.nn as nn
import torch.nn.functional as F
class Model(nn.Conv2d):
def __init__(self, in_chan, out_chan, kernel_size=3, stride=1, padding=
1, dilation=1, groups=1, bias=False, act=nn.ReLU(inplace=True), K=4,
temperature=30, temp_anneal_steps=3000):
super().__init__(in_chan, out_chan * K, kernel_size,
stride=stride, padding=padding, dilation=dilation, groups=
groups, bias=bias)
assert in_chan // 4 > 0
self.K = K
self.act = act
self.se_conv1 = nn.Conv2d(in_chan, in_chan // 4, 1, 1, 0, bias=True)
self.se_conv2 = nn.Conv2d(in_chan // 4, K, 1, 1, 0, bias=True)
self.temperature = temperature
self.temp_anneal_steps = temp_anneal_steps
self.temp_interval = (temperature - 1) / temp_anneal_steps
def get_atten(self, x):
bs, _, _h, _w = x.size()
atten = torch.mean(x, dim=(2, 3), keepdim=True)
atten = self.se_conv1(atten)
atten = self.act(atten)
atten = self.se_conv2(atten)
if self.training and self.temp_anneal_steps > 0:
atten = atten / self.temperature
self.temperature -= self.temp_interval
self.temp_anneal_steps -= 1
atten = atten.softmax(dim=1).view(bs, -1)
return atten
def forward(self, x):
bs, _, h, w = x.size()
atten = self.get_atten(x)
_out_chan, in_chan, k1, k2 = self.weight.size()
W = self.weight.view(1, self.K, -1, in_chan, k1, k2)
W = (W * atten.view(bs, self.K, 1, 1, 1, 1)).sum(dim=1)
W = W.view(-1, in_chan, k1, k2)
b = self.bias
if b is not None:
b = b.view(1, self.K, -1)
b = (b * atten.view(bs, self.K, 1)).sum(dim=1).view(-1)
x = x.view(1, -1, h, w)
out = F.conv2d(x, W, b, self.stride, self.padding, self.dilation,
self.groups * bs)
out = out.view(bs, -1, out.size(2), out.size(3))
return out
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4, 4]
|
Conv2dSame | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/xs/cxs2a7zwcw5yxvn445xldhvii7772mtsthpxnfawxoahvyf3vtaj.py
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.constant_pad_nd]
# Source node to ATen node mapping:
# x => constant_pad_nd
# Graph fragment:
# %constant_pad_nd : [num_users=2] = call_function[target=torch.ops.aten.constant_pad_nd.default](args = (%primals_3, [1, 2, 1, 2], 0.0), kwargs = {})
triton_poi_fused_constant_pad_nd_0 = async_compile.triton('triton_poi_fused_constant_pad_nd_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[1024],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_constant_pad_nd_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_constant_pad_nd_0(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 784
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = (xindex // 7) % 7
x0 = xindex % 7
x2 = (xindex // 49)
x4 = xindex
tmp0 = (-1) + x1
tmp1 = tl.full([1], 0, tl.int64)
tmp2 = tmp0 >= tmp1
tmp3 = tl.full([1], 4, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = (-1) + x0
tmp6 = tmp5 >= tmp1
tmp7 = tmp5 < tmp3
tmp8 = tmp2 & tmp4
tmp9 = tmp8 & tmp6
tmp10 = tmp9 & tmp7
tmp11 = tl.load(in_ptr0 + ((-5) + x0 + (4*x1) + (16*x2)), tmp10 & xmask, other=0.0)
tl.store(out_ptr0 + (x4), tmp11, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/32/c32v7egt4mupqssam3gmac2qgv3ujprjybthsgweflmot256qqw7.py
# Topologically Sorted Source Nodes: [conv2d], Original ATen: [aten.convolution]
# Source node to ATen node mapping:
# conv2d => convolution
# Graph fragment:
# %convolution : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%constant_pad_nd, %primals_1, %primals_2, [1, 1], [0, 0], [1, 1], False, [0, 0], 1), kwargs = {})
triton_poi_fused_convolution_1 = async_compile.triton('triton_poi_fused_convolution_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_1', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_1(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = (xindex // 16) % 4
tmp0 = tl.load(in_out_ptr0 + (x3), xmask)
tmp1 = tl.load(in_ptr0 + (x1), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + (x3), tmp2, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (4, ), (1, ))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 7, 7), (196, 49, 7, 1), torch.float32)
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.constant_pad_nd]
stream0 = get_raw_stream(0)
triton_poi_fused_constant_pad_nd_0.run(primals_3, buf0, 784, grid=grid(784), stream=stream0)
del primals_3
# Topologically Sorted Source Nodes: [conv2d], Original ATen: [aten.convolution]
buf1 = extern_kernels.convolution(buf0, primals_1, stride=(1, 1), padding=(0, 0), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf1, (4, 4, 4, 4), (64, 16, 4, 1))
buf2 = buf1; del buf1 # reuse
# Topologically Sorted Source Nodes: [conv2d], Original ATen: [aten.convolution]
triton_poi_fused_convolution_1.run(buf2, primals_2, 256, grid=grid(256), stream=stream0)
del primals_2
return (buf2, primals_1, buf0, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import math
import torch
import torch.nn.functional as F
import torch.nn as nn
from typing import List
from typing import Optional
from typing import Tuple
from torch.jit.annotations import List
def get_same_padding(x: 'int', k: 'int', s: 'int', d: 'int'):
return max((math.ceil(x / s) - 1) * s + (k - 1) * d + 1 - x, 0)
def pad_same(x, k: 'List[int]', s: 'List[int]', d: 'List[int]'=(1, 1),
value: 'float'=0):
ih, iw = x.size()[-2:]
pad_h, pad_w = get_same_padding(ih, k[0], s[0], d[0]), get_same_padding(iw,
k[1], s[1], d[1])
if pad_h > 0 or pad_w > 0:
x = F.pad(x, [pad_w // 2, pad_w - pad_w // 2, pad_h // 2, pad_h -
pad_h // 2], value=value)
return x
def conv2d_same(x, weight: 'torch.Tensor', bias: 'Optional[torch.Tensor]'=
None, stride: 'Tuple[int, int]'=(1, 1), padding: 'Tuple[int, int]'=(0,
0), dilation: 'Tuple[int, int]'=(1, 1), groups: 'int'=1):
x = pad_same(x, weight.shape[-2:], stride, dilation)
return F.conv2d(x, weight, bias, stride, (0, 0), dilation, groups)
class Conv2dSame(nn.Conv2d):
""" Tensorflow like 'SAME' convolution wrapper for 2D convolutions
"""
def __init__(self, in_channels, out_channels, kernel_size, stride=1,
padding=0, dilation=1, groups=1, bias=True):
super(Conv2dSame, self).__init__(in_channels, out_channels,
kernel_size, stride, 0, dilation, groups, bias)
def forward(self, x):
return conv2d_same(x, self.weight, self.bias, self.stride, self.
padding, self.dilation, self.groups)
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'in_channels': 4, 'out_channels': 4, 'kernel_size': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
import math
import torch.nn.functional as F
import torch.nn as nn
from typing import List
from typing import Optional
from typing import Tuple
from torch.jit.annotations import List
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_poi_fused_constant_pad_nd_0(in_ptr0, out_ptr0, xnumel, XBLOCK:
tl.constexpr):
xnumel = 784
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = xindex // 7 % 7
x0 = xindex % 7
x2 = xindex // 49
x4 = xindex
tmp0 = -1 + x1
tmp1 = tl.full([1], 0, tl.int64)
tmp2 = tmp0 >= tmp1
tmp3 = tl.full([1], 4, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = -1 + x0
tmp6 = tmp5 >= tmp1
tmp7 = tmp5 < tmp3
tmp8 = tmp2 & tmp4
tmp9 = tmp8 & tmp6
tmp10 = tmp9 & tmp7
tmp11 = tl.load(in_ptr0 + (-5 + x0 + 4 * x1 + 16 * x2), tmp10 & xmask,
other=0.0)
tl.store(out_ptr0 + x4, tmp11, xmask)
@triton.jit
def triton_poi_fused_convolution_1(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl
.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = xindex // 16 % 4
tmp0 = tl.load(in_out_ptr0 + x3, xmask)
tmp1 = tl.load(in_ptr0 + x1, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + x3, tmp2, xmask)
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (4,), (1,))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 7, 7), (196, 49, 7, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_constant_pad_nd_0[grid(784)](primals_3, buf0, 784,
XBLOCK=128, num_warps=4, num_stages=1)
del primals_3
buf1 = extern_kernels.convolution(buf0, primals_1, stride=(1, 1),
padding=(0, 0), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf1, (4, 4, 4, 4), (64, 16, 4, 1))
buf2 = buf1
del buf1
triton_poi_fused_convolution_1[grid(256)](buf2, primals_2, 256,
XBLOCK=256, num_warps=4, num_stages=1)
del primals_2
return buf2, primals_1, buf0
def get_same_padding(x: 'int', k: 'int', s: 'int', d: 'int'):
return max((math.ceil(x / s) - 1) * s + (k - 1) * d + 1 - x, 0)
def pad_same(x, k: 'List[int]', s: 'List[int]', d: 'List[int]'=(1, 1),
value: 'float'=0):
ih, iw = x.size()[-2:]
pad_h, pad_w = get_same_padding(ih, k[0], s[0], d[0]), get_same_padding(iw,
k[1], s[1], d[1])
if pad_h > 0 or pad_w > 0:
x = F.pad(x, [pad_w // 2, pad_w - pad_w // 2, pad_h // 2, pad_h -
pad_h // 2], value=value)
return x
def conv2d_same(x, weight: 'torch.Tensor', bias: 'Optional[torch.Tensor]'=
None, stride: 'Tuple[int, int]'=(1, 1), padding: 'Tuple[int, int]'=(0,
0), dilation: 'Tuple[int, int]'=(1, 1), groups: 'int'=1):
x = pad_same(x, weight.shape[-2:], stride, dilation)
return F.conv2d(x, weight, bias, stride, (0, 0), dilation, groups)
class Conv2dSameNew(nn.Conv2d):
""" Tensorflow like 'SAME' convolution wrapper for 2D convolutions
"""
def __init__(self, in_channels, out_channels, kernel_size, stride=1,
padding=0, dilation=1, groups=1, bias=True):
super(Conv2dSameNew, self).__init__(in_channels, out_channels,
kernel_size, stride, 0, dilation, groups, bias)
def forward(self, input_0):
primals_1 = self.weight
primals_2 = self.bias
primals_3 = input_0
output = call([primals_1, primals_2, primals_3])
return output[0]
| infomon/meta_nas | Conv2dSame | false | 6,883 | [
"Apache-2.0"
] | 1 | b81b7de86d26ae1ec0d6646b4277f3c918e5e35d | https://github.com/infomon/meta_nas/tree/b81b7de86d26ae1ec0d6646b4277f3c918e5e35d | import math
import torch
import torch.nn.functional as F
import torch.nn as nn
from typing import List
from typing import Optional
from typing import Tuple
from torch.jit.annotations import List
def get_same_padding(x: 'int', k: 'int', s: 'int', d: 'int'):
return max((math.ceil(x / s) - 1) * s + (k - 1) * d + 1 - x, 0)
def pad_same(x, k: 'List[int]', s: 'List[int]', d: 'List[int]'=(1, 1),
value: 'float'=0):
ih, iw = x.size()[-2:]
pad_h, pad_w = get_same_padding(ih, k[0], s[0], d[0]), get_same_padding(iw,
k[1], s[1], d[1])
if pad_h > 0 or pad_w > 0:
x = F.pad(x, [pad_w // 2, pad_w - pad_w // 2, pad_h // 2, pad_h -
pad_h // 2], value=value)
return x
def conv2d_same(x, weight: 'torch.Tensor', bias: 'Optional[torch.Tensor]'=
None, stride: 'Tuple[int, int]'=(1, 1), padding: 'Tuple[int, int]'=(0,
0), dilation: 'Tuple[int, int]'=(1, 1), groups: 'int'=1):
x = pad_same(x, weight.shape[-2:], stride, dilation)
return F.conv2d(x, weight, bias, stride, (0, 0), dilation, groups)
class Model(nn.Conv2d):
""" Tensorflow like 'SAME' convolution wrapper for 2D convolutions
"""
def __init__(self, in_channels, out_channels, kernel_size, stride=1,
padding=0, dilation=1, groups=1, bias=True):
super().__init__(in_channels, out_channels,
kernel_size, stride, 0, dilation, groups, bias)
def forward(self, x):
return conv2d_same(x, self.weight, self.bias, self.stride, self.
padding, self.dilation, self.groups)
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4, 4, 4]
|
SelfAttention_naive | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/fz/cfzmg4qtw6jgry4nhlwopodzjz62ll3n3ykfox77hwd2crdnlh2w.py
# Topologically Sorted Source Nodes: [p_attn], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# p_attn => exp
# Graph fragment:
# %mul_tensor : [num_users=2] = call_function[target=torch.ops.aten.mul.Tensor](args = (%bmm, 1), kwargs = {})
# %amax_default : [num_users=1] = call_function[target=torch.ops.aten.amax.default](args = (%mul_tensor, [2], True), kwargs = {})
# %sub_tensor : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%mul_tensor, %amax_default), kwargs = {})
# %div_tensor : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%sub_tensor, 2.0), kwargs = {})
# %exp : [num_users=2] = call_function[target=torch.ops.aten.exp.default](args = (%div_tensor,), kwargs = {})
triton_poi_fused__softmax_0 = async_compile.triton('triton_poi_fused__softmax_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_0(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp3 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp8 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp11 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp1 = 1.0
tmp2 = tmp0 * tmp1
tmp4 = tmp3 * tmp1
tmp6 = tmp5 * tmp1
tmp7 = triton_helpers.maximum(tmp4, tmp6)
tmp9 = tmp8 * tmp1
tmp10 = triton_helpers.maximum(tmp7, tmp9)
tmp12 = tmp11 * tmp1
tmp13 = triton_helpers.maximum(tmp10, tmp12)
tmp14 = tmp2 - tmp13
tmp15 = 0.5
tmp16 = tmp14 * tmp15
tmp17 = tl_math.exp(tmp16)
tl.store(out_ptr0 + (x2), tmp17, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/kj/ckjtlefzavjukjsytvkak6ek26zmzexpcbnlwelx4k5kascjxlf3.py
# Topologically Sorted Source Nodes: [p_attn], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# p_attn => div_1, sum_1
# Graph fragment:
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%exp, [2], True), kwargs = {})
# %div_1 : [num_users=2] = call_function[target=torch.ops.aten.div.Tensor](args = (%exp, %sum_1), kwargs = {})
triton_poi_fused__softmax_1 = async_compile.triton('triton_poi_fused__softmax_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_1(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + (x2), tmp8, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_2, (4, 4), (4, 1))
assert_size_stride(primals_3, (4, ), (1, ))
assert_size_stride(primals_4, (4, 4), (4, 1))
assert_size_stride(primals_5, (4, ), (1, ))
assert_size_stride(primals_6, (4, 4), (4, 1))
assert_size_stride(primals_7, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [queries], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_3, reinterpret_tensor(primals_1, (16, 4), (4, 1), 0), reinterpret_tensor(primals_2, (4, 4), (1, 4), 0), alpha=1, beta=1, out=buf0)
del primals_2
del primals_3
buf1 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [keys], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_5, reinterpret_tensor(primals_1, (16, 4), (4, 1), 0), reinterpret_tensor(primals_4, (4, 4), (1, 4), 0), alpha=1, beta=1, out=buf1)
del primals_4
del primals_5
buf2 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [values], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_7, reinterpret_tensor(primals_1, (16, 4), (4, 1), 0), reinterpret_tensor(primals_6, (4, 4), (1, 4), 0), alpha=1, beta=1, out=buf2)
del primals_6
del primals_7
buf3 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [matmul], Original ATen: [aten.bmm]
extern_kernels.bmm(reinterpret_tensor(buf0, (4, 4, 4), (16, 4, 1), 0), reinterpret_tensor(buf1, (4, 4, 4), (16, 1, 4), 0), out=buf3)
buf4 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [p_attn], Original ATen: [aten._softmax]
stream0 = get_raw_stream(0)
triton_poi_fused__softmax_0.run(buf3, buf4, 64, grid=grid(64), stream=stream0)
buf5 = buf3; del buf3 # reuse
# Topologically Sorted Source Nodes: [p_attn], Original ATen: [aten._softmax]
triton_poi_fused__softmax_1.run(buf4, buf5, 64, grid=grid(64), stream=stream0)
buf6 = buf4; del buf4 # reuse
# Topologically Sorted Source Nodes: [z], Original ATen: [aten.bmm]
extern_kernels.bmm(buf5, reinterpret_tensor(buf2, (4, 4, 4), (16, 4, 1), 0), out=buf6)
return (buf6, reinterpret_tensor(primals_1, (16, 4), (4, 1), 0), buf5, reinterpret_tensor(buf2, (4, 4, 4), (16, 1, 4), 0), reinterpret_tensor(buf0, (4, 4, 4), (16, 1, 4), 0), reinterpret_tensor(buf1, (4, 4, 4), (16, 4, 1), 0), )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import math
import torch
from torch import nn
import torch.nn.functional as F
class SelfAttention_naive(nn.Module):
def __init__(self, dim_emb, dim_internal, heads=8, mask=False, dropout=
0.0, dtype=torch.float32):
"""
A single self attention block
:param dim_emb: embedding dimension
:param dim_internal: dimension of internal representation, usually the same as dim_emb
:param head: number of multi head
:param mask
"""
super().__init__()
self.dim_emb = dim_emb
self.dim_internal = dim_internal
self.heads = heads
self.mask = mask
self.toqueries = nn.Linear(dim_emb, dim_internal).type(dtype)
self.tokeys = nn.Linear(dim_emb, dim_internal).type(dtype)
self.tovalues = nn.Linear(dim_emb, dim_internal).type(dtype)
self.kSqrt_dim_emb = math.sqrt(self.dim_emb)
def forward(self, x):
_b, _t, e = x.size()
assert e == self.dim_emb, f'Input embedding ({e}) should match the layer embedding ({self.dim_emb})'
queries = self.toqueries(x)
keys = self.tokeys(x)
values = self.tovalues(x)
keys_transposed = keys.transpose(-2, -1)
dot = torch.matmul(queries, keys_transposed) / self.kSqrt_dim_emb
p_attn = F.softmax(dot, dim=2)
z = torch.matmul(p_attn, values)
return z
def get_inputs():
return [torch.rand([4, 4, 4])]
def get_init_inputs():
return [[], {'dim_emb': 4, 'dim_internal': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import math as tl_math
import math
from torch import nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused__softmax_0(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp3 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp8 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp11 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last'
)
tmp1 = 1.0
tmp2 = tmp0 * tmp1
tmp4 = tmp3 * tmp1
tmp6 = tmp5 * tmp1
tmp7 = triton_helpers.maximum(tmp4, tmp6)
tmp9 = tmp8 * tmp1
tmp10 = triton_helpers.maximum(tmp7, tmp9)
tmp12 = tmp11 * tmp1
tmp13 = triton_helpers.maximum(tmp10, tmp12)
tmp14 = tmp2 - tmp13
tmp15 = 0.5
tmp16 = tmp14 * tmp15
tmp17 = tl_math.exp(tmp16)
tl.store(out_ptr0 + x2, tmp17, xmask)
@triton.jit
def triton_poi_fused__softmax_1(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + x2, tmp8, xmask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7) = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_2, (4, 4), (4, 1))
assert_size_stride(primals_3, (4,), (1,))
assert_size_stride(primals_4, (4, 4), (4, 1))
assert_size_stride(primals_5, (4,), (1,))
assert_size_stride(primals_6, (4, 4), (4, 1))
assert_size_stride(primals_7, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.addmm(primals_3, reinterpret_tensor(primals_1, (16,
4), (4, 1), 0), reinterpret_tensor(primals_2, (4, 4), (1, 4), 0
), alpha=1, beta=1, out=buf0)
del primals_2
del primals_3
buf1 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.addmm(primals_5, reinterpret_tensor(primals_1, (16,
4), (4, 1), 0), reinterpret_tensor(primals_4, (4, 4), (1, 4), 0
), alpha=1, beta=1, out=buf1)
del primals_4
del primals_5
buf2 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.addmm(primals_7, reinterpret_tensor(primals_1, (16,
4), (4, 1), 0), reinterpret_tensor(primals_6, (4, 4), (1, 4), 0
), alpha=1, beta=1, out=buf2)
del primals_6
del primals_7
buf3 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
extern_kernels.bmm(reinterpret_tensor(buf0, (4, 4, 4), (16, 4, 1),
0), reinterpret_tensor(buf1, (4, 4, 4), (16, 1, 4), 0), out=buf3)
buf4 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused__softmax_0[grid(64)](buf3, buf4, 64, XBLOCK=64,
num_warps=1, num_stages=1)
buf5 = buf3
del buf3
triton_poi_fused__softmax_1[grid(64)](buf4, buf5, 64, XBLOCK=64,
num_warps=1, num_stages=1)
buf6 = buf4
del buf4
extern_kernels.bmm(buf5, reinterpret_tensor(buf2, (4, 4, 4), (16, 4,
1), 0), out=buf6)
return buf6, reinterpret_tensor(primals_1, (16, 4), (4, 1), 0
), buf5, reinterpret_tensor(buf2, (4, 4, 4), (16, 1, 4), 0
), reinterpret_tensor(buf0, (4, 4, 4), (16, 1, 4), 0
), reinterpret_tensor(buf1, (4, 4, 4), (16, 4, 1), 0)
class SelfAttention_naiveNew(nn.Module):
def __init__(self, dim_emb, dim_internal, heads=8, mask=False, dropout=
0.0, dtype=torch.float32):
"""
A single self attention block
:param dim_emb: embedding dimension
:param dim_internal: dimension of internal representation, usually the same as dim_emb
:param head: number of multi head
:param mask
"""
super().__init__()
self.dim_emb = dim_emb
self.dim_internal = dim_internal
self.heads = heads
self.mask = mask
self.toqueries = nn.Linear(dim_emb, dim_internal).type(dtype)
self.tokeys = nn.Linear(dim_emb, dim_internal).type(dtype)
self.tovalues = nn.Linear(dim_emb, dim_internal).type(dtype)
self.kSqrt_dim_emb = math.sqrt(self.dim_emb)
def forward(self, input_0):
primals_2 = self.toqueries.weight
primals_3 = self.toqueries.bias
primals_4 = self.tokeys.weight
primals_5 = self.tokeys.bias
primals_6 = self.tovalues.weight
primals_7 = self.tovalues.bias
primals_1 = input_0
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7])
return output[0]
| insop/transformer_simple | SelfAttention_naive | false | 6,884 | [
"Apache-2.0"
] | 1 | d07e6c3b9ddc9687d332ac3a980bbce22880ad46 | https://github.com/insop/transformer_simple/tree/d07e6c3b9ddc9687d332ac3a980bbce22880ad46 | import math
import torch
from torch import nn
import torch.nn.functional as F
class Model(nn.Module):
def __init__(self, dim_emb, dim_internal, heads=8, mask=False, dropout=
0.0, dtype=torch.float32):
"""
A single self attention block
:param dim_emb: embedding dimension
:param dim_internal: dimension of internal representation, usually the same as dim_emb
:param head: number of multi head
:param mask
"""
super().__init__()
self.dim_emb = dim_emb
self.dim_internal = dim_internal
self.heads = heads
self.mask = mask
self.toqueries = nn.Linear(dim_emb, dim_internal).type(dtype)
self.tokeys = nn.Linear(dim_emb, dim_internal).type(dtype)
self.tovalues = nn.Linear(dim_emb, dim_internal).type(dtype)
self.kSqrt_dim_emb = math.sqrt(self.dim_emb)
def forward(self, x):
_b, _t, e = x.size()
assert e == self.dim_emb, f'Input embedding ({e}) should match the layer embedding ({self.dim_emb})'
queries = self.toqueries(x)
keys = self.tokeys(x)
values = self.tovalues(x)
keys_transposed = keys.transpose(-2, -1)
dot = torch.matmul(queries, keys_transposed) / self.kSqrt_dim_emb
p_attn = F.softmax(dot, dim=2)
z = torch.matmul(p_attn, values)
return z
def get_inputs():
return [torch.rand([4, 4, 4])]
def get_init_inputs():
return [4, 4]
|
MultiHead | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/p6/cp62j5tv2tk4ynnj52fvxnzwu3qhxoiovlhltiksafkjm6cs5wpz.py
# Topologically Sorted Source Nodes: [QWs], Original ATen: [aten.cat]
# Source node to ATen node mapping:
# QWs => cat
# Graph fragment:
# %cat : [num_users=2] = call_function[target=torch.ops.aten.cat.default](args = ([%getitem, %getitem_1, %getitem_2, %getitem_3],), kwargs = {})
triton_poi_fused_cat_0 = async_compile.triton('triton_poi_fused_cat_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_cat_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 4, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_cat_0(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = (xindex // 4)
x0 = xindex % 4
x2 = xindex
tmp0 = x1
tmp1 = tl.full([1], 0, tl.int64)
tmp2 = tmp0 >= tmp1
tmp3 = tl.full([1], 4, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = tl.load(in_ptr0 + ((4*x0) + (16*x1)), tmp4 & xmask, eviction_policy='evict_last', other=0.0)
tmp6 = tmp0 >= tmp3
tmp7 = tl.full([1], 8, tl.int64)
tmp8 = tmp0 < tmp7
tmp9 = tmp6 & tmp8
tmp10 = tl.load(in_ptr0 + (1 + (4*x0) + (16*((-4) + x1))), tmp9 & xmask, eviction_policy='evict_last', other=0.0)
tmp11 = tmp0 >= tmp7
tmp12 = tl.full([1], 12, tl.int64)
tmp13 = tmp0 < tmp12
tmp14 = tmp11 & tmp13
tmp15 = tl.load(in_ptr0 + (2 + (4*x0) + (16*((-8) + x1))), tmp14 & xmask, eviction_policy='evict_last', other=0.0)
tmp16 = tmp0 >= tmp12
tmp17 = tl.full([1], 16, tl.int64)
tmp18 = tmp0 < tmp17
tmp19 = tl.load(in_ptr0 + (3 + (4*x0) + (16*((-12) + x1))), tmp16 & xmask, eviction_policy='evict_last', other=0.0)
tmp20 = tl.where(tmp14, tmp15, tmp19)
tmp21 = tl.where(tmp9, tmp10, tmp20)
tmp22 = tl.where(tmp4, tmp5, tmp21)
tl.store(out_ptr0 + (x2), tmp22, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/s2/cs2rk3o3kmhydx4oijp6rsdb5atcrq5axy4adadrpl7gkt7scies.py
# Topologically Sorted Source Nodes: [w_1], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# w_1 => exp
# Graph fragment:
# %mul_tensor : [num_users=2] = call_function[target=torch.ops.aten.mul.Tensor](args = (%bmm, 1), kwargs = {})
# %amax_default : [num_users=1] = call_function[target=torch.ops.aten.amax.default](args = (%mul_tensor, [-1], True), kwargs = {})
# %sub_tensor : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%mul_tensor, %amax_default), kwargs = {})
# %div_tensor : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%sub_tensor, 1.0), kwargs = {})
# %exp : [num_users=2] = call_function[target=torch.ops.aten.exp.default](args = (%div_tensor,), kwargs = {})
triton_poi_fused__softmax_1 = async_compile.triton('triton_poi_fused__softmax_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_1(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp3 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp8 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp11 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp1 = 1.0
tmp2 = tmp0 * tmp1
tmp4 = tmp3 * tmp1
tmp6 = tmp5 * tmp1
tmp7 = triton_helpers.maximum(tmp4, tmp6)
tmp9 = tmp8 * tmp1
tmp10 = triton_helpers.maximum(tmp7, tmp9)
tmp12 = tmp11 * tmp1
tmp13 = triton_helpers.maximum(tmp10, tmp12)
tmp14 = tmp2 - tmp13
tmp15 = tmp14 * tmp1
tmp16 = tl_math.exp(tmp15)
tl.store(out_ptr0 + (x2), tmp16, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/3f/c3fx6bzkalkw7u7askqdnz4rzlcoyqiec4r434sjc5x3axxgkrmr.py
# Topologically Sorted Source Nodes: [w_1], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# w_1 => div_1, sum_1
# Graph fragment:
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%exp, [-1], True), kwargs = {})
# %div_1 : [num_users=2] = call_function[target=torch.ops.aten.div.Tensor](args = (%exp, %sum_1), kwargs = {})
triton_poi_fused__softmax_2 = async_compile.triton('triton_poi_fused__softmax_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_2(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + (x2), tmp8, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/7l/c7l66q7o72hyct67pldiidtrovmmlldouzc4gdlmwgzljxbtlzxy.py
# Topologically Sorted Source Nodes: [cat_3], Original ATen: [aten.cat]
# Source node to ATen node mapping:
# cat_3 => cat_3
# Graph fragment:
# %cat_3 : [num_users=1] = call_function[target=torch.ops.aten.cat.default](args = ([%getitem_12, %getitem_13, %getitem_14, %getitem_15], -1), kwargs = {})
triton_poi_fused_cat_3 = async_compile.triton('triton_poi_fused_cat_3', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_cat_3', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 4, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_cat_3(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 4
x1 = (xindex // 4)
x2 = xindex
tmp0 = x0
tmp1 = tl.full([1], 0, tl.int64)
tmp2 = tmp0 >= tmp1
tmp3 = tl.full([1], 1, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = tl.load(in_ptr0 + (x1), tmp4 & xmask, eviction_policy='evict_last', other=0.0)
tmp6 = tmp0 >= tmp3
tmp7 = tl.full([1], 2, tl.int64)
tmp8 = tmp0 < tmp7
tmp9 = tmp6 & tmp8
tmp10 = tl.load(in_ptr0 + (16 + x1), tmp9 & xmask, eviction_policy='evict_last', other=0.0)
tmp11 = tmp0 >= tmp7
tmp12 = tl.full([1], 3, tl.int64)
tmp13 = tmp0 < tmp12
tmp14 = tmp11 & tmp13
tmp15 = tl.load(in_ptr0 + (32 + x1), tmp14 & xmask, eviction_policy='evict_last', other=0.0)
tmp16 = tmp0 >= tmp12
tmp17 = tl.full([1], 4, tl.int64)
tmp18 = tmp0 < tmp17
tmp19 = tl.load(in_ptr0 + (48 + x1), tmp16 & xmask, eviction_policy='evict_last', other=0.0)
tmp20 = tl.where(tmp14, tmp15, tmp19)
tmp21 = tl.where(tmp9, tmp10, tmp20)
tmp22 = tl.where(tmp4, tmp5, tmp21)
tl.store(out_ptr0 + (x2), tmp22, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7 = args
args.clear()
assert_size_stride(primals_1, (4, 4), (4, 1))
assert_size_stride(primals_2, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_3, (4, 4), (4, 1))
assert_size_stride(primals_4, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_5, (4, 4), (4, 1))
assert_size_stride(primals_6, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_7, (4, 4), (4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [linear], Original ATen: [aten.mm]
extern_kernels.mm(reinterpret_tensor(primals_2, (16, 4), (4, 1), 0), reinterpret_tensor(primals_1, (4, 4), (1, 4), 0), out=buf0)
del primals_1
buf1 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [linear_1], Original ATen: [aten.mm]
extern_kernels.mm(reinterpret_tensor(primals_4, (16, 4), (4, 1), 0), reinterpret_tensor(primals_3, (4, 4), (1, 4), 0), out=buf1)
del primals_3
buf2 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [linear_2], Original ATen: [aten.mm]
extern_kernels.mm(reinterpret_tensor(primals_6, (16, 4), (4, 1), 0), reinterpret_tensor(primals_5, (4, 4), (1, 4), 0), out=buf2)
del primals_5
buf3 = empty_strided_cuda((16, 4, 1), (4, 1, 64), torch.float32)
# Topologically Sorted Source Nodes: [QWs], Original ATen: [aten.cat]
stream0 = get_raw_stream(0)
triton_poi_fused_cat_0.run(buf0, buf3, 64, grid=grid(64), stream=stream0)
buf4 = reinterpret_tensor(buf0, (16, 4, 1), (4, 1, 64), 0); del buf0 # reuse
# Topologically Sorted Source Nodes: [VWs], Original ATen: [aten.cat]
triton_poi_fused_cat_0.run(buf2, buf4, 64, grid=grid(64), stream=stream0)
buf5 = reinterpret_tensor(buf2, (16, 4, 1), (4, 1, 1), 0); del buf2 # reuse
# Topologically Sorted Source Nodes: [KWs], Original ATen: [aten.cat]
triton_poi_fused_cat_0.run(buf1, buf5, 64, grid=grid(64), stream=stream0)
buf6 = empty_strided_cuda((16, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [w], Original ATen: [aten.bmm]
extern_kernels.bmm(buf3, reinterpret_tensor(buf5, (16, 1, 4), (4, 0, 1), 0), out=buf6)
buf7 = empty_strided_cuda((16, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [w_1], Original ATen: [aten._softmax]
triton_poi_fused__softmax_1.run(buf6, buf7, 256, grid=grid(256), stream=stream0)
buf8 = buf6; del buf6 # reuse
# Topologically Sorted Source Nodes: [w_1], Original ATen: [aten._softmax]
triton_poi_fused__softmax_2.run(buf7, buf8, 256, grid=grid(256), stream=stream0)
del buf7
buf9 = reinterpret_tensor(buf1, (16, 4, 1), (4, 1, 1), 0); del buf1 # reuse
# Topologically Sorted Source Nodes: [c], Original ATen: [aten.bmm]
extern_kernels.bmm(buf8, buf4, out=buf9)
buf10 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [cat_3], Original ATen: [aten.cat]
triton_poi_fused_cat_3.run(buf9, buf10, 64, grid=grid(64), stream=stream0)
buf11 = reinterpret_tensor(buf9, (16, 4), (4, 1), 0); del buf9 # reuse
# Topologically Sorted Source Nodes: [c_1], Original ATen: [aten.mm]
extern_kernels.mm(reinterpret_tensor(buf10, (16, 4), (4, 1), 0), reinterpret_tensor(primals_7, (4, 4), (1, 4), 0), out=buf11)
return (reinterpret_tensor(buf11, (4, 4, 4), (16, 4, 1), 0), reinterpret_tensor(primals_2, (16, 4), (4, 1), 0), reinterpret_tensor(primals_4, (16, 4), (4, 1), 0), reinterpret_tensor(primals_6, (16, 4), (4, 1), 0), buf8, reinterpret_tensor(buf10, (16, 4), (4, 1), 0), primals_7, reinterpret_tensor(buf4, (16, 1, 4), (4, 1, 1), 0), reinterpret_tensor(buf3, (16, 1, 4), (4, 1, 1), 0), buf5, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
class Attention(nn.Module):
def __init__(self):
super().__init__()
self.softmax = nn.Softmax(dim=-1)
def forward(self, Q, K, V, mask=None, dk=64):
w = torch.bmm(Q, K.transpose(1, 2))
if mask is not None:
assert w.size() == mask.size()
w.masked_fill_(mask, -float('inf'))
w = self.softmax(w / dk ** 0.5)
c = torch.bmm(w, V)
return c
class MultiHead(nn.Module):
def __init__(self, hidden_size, n_splits):
super().__init__()
self.hidden_size = hidden_size
self.n_splits = n_splits
self.Q_linear = nn.Linear(hidden_size, hidden_size, bias=False)
self.K_linear = nn.Linear(hidden_size, hidden_size, bias=False)
self.V_linear = nn.Linear(hidden_size, hidden_size, bias=False)
self.linear = nn.Linear(hidden_size, hidden_size, bias=False)
self.attn = Attention()
def forward(self, Q, K, V, mask=None):
QWs = self.Q_linear(Q).split(self.hidden_size // self.n_splits, dim=-1)
KWs = self.K_linear(K).split(self.hidden_size // self.n_splits, dim=-1)
VWs = self.V_linear(V).split(self.hidden_size // self.n_splits, dim=-1)
QWs = torch.cat(QWs, dim=0)
KWs = torch.cat(KWs, dim=0)
VWs = torch.cat(VWs, dim=0)
if mask is not None:
mask = torch.cat([mask for _ in range(self.n_splits)], dim=0)
c = self.attn(QWs, KWs, VWs, mask=mask, dk=self.hidden_size // self
.n_splits)
c = c.split(Q.size(0), dim=0)
c = self.linear(torch.cat(c, dim=-1))
return c
def get_inputs():
return [torch.rand([4, 4, 4]), torch.rand([4, 4, 4]), torch.rand([4, 4, 4])
]
def get_init_inputs():
return [[], {'hidden_size': 4, 'n_splits': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import math as tl_math
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_cat_0(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = xindex // 4
x0 = xindex % 4
x2 = xindex
tmp0 = x1
tl.full([1], 0, tl.int64)
tmp3 = tl.full([1], 4, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = tl.load(in_ptr0 + (4 * x0 + 16 * x1), tmp4 & xmask,
eviction_policy='evict_last', other=0.0)
tmp6 = tmp0 >= tmp3
tmp7 = tl.full([1], 8, tl.int64)
tmp8 = tmp0 < tmp7
tmp9 = tmp6 & tmp8
tmp10 = tl.load(in_ptr0 + (1 + 4 * x0 + 16 * (-4 + x1)), tmp9 & xmask,
eviction_policy='evict_last', other=0.0)
tmp11 = tmp0 >= tmp7
tmp12 = tl.full([1], 12, tl.int64)
tmp13 = tmp0 < tmp12
tmp14 = tmp11 & tmp13
tmp15 = tl.load(in_ptr0 + (2 + 4 * x0 + 16 * (-8 + x1)), tmp14 & xmask,
eviction_policy='evict_last', other=0.0)
tmp16 = tmp0 >= tmp12
tl.full([1], 16, tl.int64)
tmp19 = tl.load(in_ptr0 + (3 + 4 * x0 + 16 * (-12 + x1)), tmp16 & xmask,
eviction_policy='evict_last', other=0.0)
tmp20 = tl.where(tmp14, tmp15, tmp19)
tmp21 = tl.where(tmp9, tmp10, tmp20)
tmp22 = tl.where(tmp4, tmp5, tmp21)
tl.store(out_ptr0 + x2, tmp22, xmask)
@triton.jit
def triton_poi_fused__softmax_1(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp3 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp8 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp11 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last'
)
tmp1 = 1.0
tmp2 = tmp0 * tmp1
tmp4 = tmp3 * tmp1
tmp6 = tmp5 * tmp1
tmp7 = triton_helpers.maximum(tmp4, tmp6)
tmp9 = tmp8 * tmp1
tmp10 = triton_helpers.maximum(tmp7, tmp9)
tmp12 = tmp11 * tmp1
tmp13 = triton_helpers.maximum(tmp10, tmp12)
tmp14 = tmp2 - tmp13
tmp15 = tmp14 * tmp1
tmp16 = tl_math.exp(tmp15)
tl.store(out_ptr0 + x2, tmp16, xmask)
@triton.jit
def triton_poi_fused__softmax_2(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + x2, tmp8, xmask)
@triton.jit
def triton_poi_fused_cat_3(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 4
x1 = xindex // 4
x2 = xindex
tmp0 = x0
tl.full([1], 0, tl.int64)
tmp3 = tl.full([1], 1, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = tl.load(in_ptr0 + x1, tmp4 & xmask, eviction_policy='evict_last',
other=0.0)
tmp6 = tmp0 >= tmp3
tmp7 = tl.full([1], 2, tl.int64)
tmp8 = tmp0 < tmp7
tmp9 = tmp6 & tmp8
tmp10 = tl.load(in_ptr0 + (16 + x1), tmp9 & xmask, eviction_policy=
'evict_last', other=0.0)
tmp11 = tmp0 >= tmp7
tmp12 = tl.full([1], 3, tl.int64)
tmp13 = tmp0 < tmp12
tmp14 = tmp11 & tmp13
tmp15 = tl.load(in_ptr0 + (32 + x1), tmp14 & xmask, eviction_policy=
'evict_last', other=0.0)
tmp16 = tmp0 >= tmp12
tl.full([1], 4, tl.int64)
tmp19 = tl.load(in_ptr0 + (48 + x1), tmp16 & xmask, eviction_policy=
'evict_last', other=0.0)
tmp20 = tl.where(tmp14, tmp15, tmp19)
tmp21 = tl.where(tmp9, tmp10, tmp20)
tmp22 = tl.where(tmp4, tmp5, tmp21)
tl.store(out_ptr0 + x2, tmp22, xmask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7) = args
args.clear()
assert_size_stride(primals_1, (4, 4), (4, 1))
assert_size_stride(primals_2, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_3, (4, 4), (4, 1))
assert_size_stride(primals_4, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_5, (4, 4), (4, 1))
assert_size_stride(primals_6, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_7, (4, 4), (4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_2, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_1, (4, 4), (1, 4), 0), out=buf0)
del primals_1
buf1 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_4, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_3, (4, 4), (1, 4), 0), out=buf1)
del primals_3
buf2 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_6, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_5, (4, 4), (1, 4), 0), out=buf2)
del primals_5
buf3 = empty_strided_cuda((16, 4, 1), (4, 1, 64), torch.float32)
get_raw_stream(0)
triton_poi_fused_cat_0[grid(64)](buf0, buf3, 64, XBLOCK=64,
num_warps=1, num_stages=1)
buf4 = reinterpret_tensor(buf0, (16, 4, 1), (4, 1, 64), 0)
del buf0
triton_poi_fused_cat_0[grid(64)](buf2, buf4, 64, XBLOCK=64,
num_warps=1, num_stages=1)
buf5 = reinterpret_tensor(buf2, (16, 4, 1), (4, 1, 1), 0)
del buf2
triton_poi_fused_cat_0[grid(64)](buf1, buf5, 64, XBLOCK=64,
num_warps=1, num_stages=1)
buf6 = empty_strided_cuda((16, 4, 4), (16, 4, 1), torch.float32)
extern_kernels.bmm(buf3, reinterpret_tensor(buf5, (16, 1, 4), (4, 0,
1), 0), out=buf6)
buf7 = empty_strided_cuda((16, 4, 4), (16, 4, 1), torch.float32)
triton_poi_fused__softmax_1[grid(256)](buf6, buf7, 256, XBLOCK=256,
num_warps=4, num_stages=1)
buf8 = buf6
del buf6
triton_poi_fused__softmax_2[grid(256)](buf7, buf8, 256, XBLOCK=256,
num_warps=4, num_stages=1)
del buf7
buf9 = reinterpret_tensor(buf1, (16, 4, 1), (4, 1, 1), 0)
del buf1
extern_kernels.bmm(buf8, buf4, out=buf9)
buf10 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
triton_poi_fused_cat_3[grid(64)](buf9, buf10, 64, XBLOCK=64,
num_warps=1, num_stages=1)
buf11 = reinterpret_tensor(buf9, (16, 4), (4, 1), 0)
del buf9
extern_kernels.mm(reinterpret_tensor(buf10, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_7, (4, 4), (1, 4), 0), out=buf11)
return reinterpret_tensor(buf11, (4, 4, 4), (16, 4, 1), 0
), reinterpret_tensor(primals_2, (16, 4), (4, 1), 0
), reinterpret_tensor(primals_4, (16, 4), (4, 1), 0
), reinterpret_tensor(primals_6, (16, 4), (4, 1), 0
), buf8, reinterpret_tensor(buf10, (16, 4), (4, 1), 0
), primals_7, reinterpret_tensor(buf4, (16, 1, 4), (4, 1, 1), 0
), reinterpret_tensor(buf3, (16, 1, 4), (4, 1, 1), 0), buf5
class Attention(nn.Module):
def __init__(self):
super().__init__()
self.softmax = nn.Softmax(dim=-1)
def forward(self, Q, K, V, mask=None, dk=64):
w = torch.bmm(Q, K.transpose(1, 2))
if mask is not None:
assert w.size() == mask.size()
w.masked_fill_(mask, -float('inf'))
w = self.softmax(w / dk ** 0.5)
c = torch.bmm(w, V)
return c
class MultiHeadNew(nn.Module):
def __init__(self, hidden_size, n_splits):
super().__init__()
self.hidden_size = hidden_size
self.n_splits = n_splits
self.Q_linear = nn.Linear(hidden_size, hidden_size, bias=False)
self.K_linear = nn.Linear(hidden_size, hidden_size, bias=False)
self.V_linear = nn.Linear(hidden_size, hidden_size, bias=False)
self.linear = nn.Linear(hidden_size, hidden_size, bias=False)
self.attn = Attention()
def forward(self, input_0, input_1, input_2):
primals_1 = self.Q_linear.weight
primals_3 = self.K_linear.weight
primals_5 = self.V_linear.weight
primals_7 = self.linear.weight
primals_2 = input_0
primals_4 = input_1
primals_6 = input_2
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7])
return output[0]
| iml1111/machine-translation | MultiHead | false | 6,885 | [
"MIT"
] | 1 | a7dd673efbe8a172c1df49e0d50482dc84008c37 | https://github.com/iml1111/machine-translation/tree/a7dd673efbe8a172c1df49e0d50482dc84008c37 | import torch
import torch.nn as nn
class Attention(nn.Module):
def __init__(self):
super().__init__()
self.softmax = nn.Softmax(dim=-1)
def forward(self, Q, K, V, mask=None, dk=64):
w = torch.bmm(Q, K.transpose(1, 2))
if mask is not None:
assert w.size() == mask.size()
w.masked_fill_(mask, -float('inf'))
w = self.softmax(w / dk ** 0.5)
c = torch.bmm(w, V)
return c
class Model(nn.Module):
def __init__(self, hidden_size, n_splits):
super().__init__()
self.hidden_size = hidden_size
self.n_splits = n_splits
self.Q_linear = nn.Linear(hidden_size, hidden_size, bias=False)
self.K_linear = nn.Linear(hidden_size, hidden_size, bias=False)
self.V_linear = nn.Linear(hidden_size, hidden_size, bias=False)
self.linear = nn.Linear(hidden_size, hidden_size, bias=False)
self.attn = Attention()
def forward(self, Q, K, V, mask=None):
QWs = self.Q_linear(Q).split(self.hidden_size // self.n_splits, dim=-1)
KWs = self.K_linear(K).split(self.hidden_size // self.n_splits, dim=-1)
VWs = self.V_linear(V).split(self.hidden_size // self.n_splits, dim=-1)
QWs = torch.cat(QWs, dim=0)
KWs = torch.cat(KWs, dim=0)
VWs = torch.cat(VWs, dim=0)
if mask is not None:
mask = torch.cat([mask for _ in range(self.n_splits)], dim=0)
c = self.attn(QWs, KWs, VWs, mask=mask, dk=self.hidden_size // self
.n_splits)
c = c.split(Q.size(0), dim=0)
c = self.linear(torch.cat(c, dim=-1))
return c
def get_inputs():
return [torch.rand([4, 4, 4]), torch.rand([4, 4, 4]), torch.rand([4, 4, 4])
]
def get_init_inputs():
return [4, 4]
|
IOU | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/32/c32cfdopzg4m774tjsidpuyrzqvzipj7lbnlkuiw6gidczl6hvux.py
# Topologically Sorted Source Nodes: [sum_2, sum_3, add, mul, Iand1, Ior1, IoU1, sub_1, IoU, sum_5, sum_6, add_2, mul_1, Iand1_1, Ior1_1, IoU1_1, sub_3, IoU_1, sum_8, sum_9, add_4, mul_2, Iand1_2, Ior1_2, IoU1_2, sub_5, IoU_2, sum_11, sum_12, add_6, mul_3, Iand1_3, Ior1_3, IoU1_3, sub_7, IoU_3, truediv_4], Original ATen: [aten.sum, aten.add, aten.mul, aten.sub, aten.div, aten.rsub]
# Source node to ATen node mapping:
# Iand1 => sum_1
# Iand1_1 => sum_4
# Iand1_2 => sum_7
# Iand1_3 => sum_10
# IoU => add_1
# IoU1 => div
# IoU1_1 => div_1
# IoU1_2 => div_2
# IoU1_3 => div_3
# IoU_1 => add_3
# IoU_2 => add_5
# IoU_3 => add_7
# Ior1 => sub
# Ior1_1 => sub_2
# Ior1_2 => sub_4
# Ior1_3 => sub_6
# add => add
# add_2 => add_2
# add_4 => add_4
# add_6 => add_6
# mul => mul
# mul_1 => mul_1
# mul_2 => mul_2
# mul_3 => mul_3
# sub_1 => sub_1
# sub_3 => sub_3
# sub_5 => sub_5
# sub_7 => sub_7
# sum_11 => sum_11
# sum_12 => sum_12
# sum_2 => sum_2
# sum_3 => sum_3
# sum_5 => sum_5
# sum_6 => sum_6
# sum_8 => sum_8
# sum_9 => sum_9
# truediv_4 => div_4
# Graph fragment:
# %sum_2 : [num_users=1] = call_function[target=torch.ops.aten.sum.default](args = (%select_2,), kwargs = {})
# %sum_3 : [num_users=1] = call_function[target=torch.ops.aten.sum.default](args = (%select_3,), kwargs = {})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%sum_2, %sum_3), kwargs = {})
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%select, %select_1), kwargs = {})
# %sum_1 : [num_users=2] = call_function[target=torch.ops.aten.sum.default](args = (%mul,), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%add, %sum_1), kwargs = {})
# %div : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%sum_1, %sub), kwargs = {})
# %sub_1 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (1, %div), kwargs = {})
# %add_1 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%sub_1, 0.0), kwargs = {})
# %sum_5 : [num_users=1] = call_function[target=torch.ops.aten.sum.default](args = (%select_6,), kwargs = {})
# %sum_6 : [num_users=1] = call_function[target=torch.ops.aten.sum.default](args = (%select_7,), kwargs = {})
# %add_2 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%sum_5, %sum_6), kwargs = {})
# %mul_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%select_4, %select_5), kwargs = {})
# %sum_4 : [num_users=2] = call_function[target=torch.ops.aten.sum.default](args = (%mul_1,), kwargs = {})
# %sub_2 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%add_2, %sum_4), kwargs = {})
# %div_1 : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%sum_4, %sub_2), kwargs = {})
# %sub_3 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (1, %div_1), kwargs = {})
# %add_3 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%add_1, %sub_3), kwargs = {})
# %sum_8 : [num_users=1] = call_function[target=torch.ops.aten.sum.default](args = (%select_10,), kwargs = {})
# %sum_9 : [num_users=1] = call_function[target=torch.ops.aten.sum.default](args = (%select_11,), kwargs = {})
# %add_4 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%sum_8, %sum_9), kwargs = {})
# %mul_2 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%select_8, %select_9), kwargs = {})
# %sum_7 : [num_users=2] = call_function[target=torch.ops.aten.sum.default](args = (%mul_2,), kwargs = {})
# %sub_4 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%add_4, %sum_7), kwargs = {})
# %div_2 : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%sum_7, %sub_4), kwargs = {})
# %sub_5 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (1, %div_2), kwargs = {})
# %add_5 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%add_3, %sub_5), kwargs = {})
# %sum_11 : [num_users=1] = call_function[target=torch.ops.aten.sum.default](args = (%select_14,), kwargs = {})
# %sum_12 : [num_users=1] = call_function[target=torch.ops.aten.sum.default](args = (%select_15,), kwargs = {})
# %add_6 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%sum_11, %sum_12), kwargs = {})
# %mul_3 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%select_12, %select_13), kwargs = {})
# %sum_10 : [num_users=2] = call_function[target=torch.ops.aten.sum.default](args = (%mul_3,), kwargs = {})
# %sub_6 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%add_6, %sum_10), kwargs = {})
# %div_3 : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%sum_10, %sub_6), kwargs = {})
# %sub_7 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (1, %div_3), kwargs = {})
# %add_7 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%add_5, %sub_7), kwargs = {})
# %div_4 : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%add_7, 4), kwargs = {})
triton_per_fused_add_div_mul_rsub_sub_sum_0 = async_compile.triton('triton_per_fused_add_div_mul_rsub_sub_sum_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[1, 64],
reduction_hint=ReductionHint.INNER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {3: 1}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 4), equal_to_1=(3,))]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused_add_div_mul_rsub_sub_sum_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 8, 'num_reduction': 12, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused_add_div_mul_rsub_sub_sum_0(in_out_ptr0, in_ptr0, in_ptr1, xnumel, rnumel, XBLOCK : tl.constexpr):
xnumel = 1
rnumel = 64
RBLOCK: tl.constexpr = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[None, :]
roffset = 0
rmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
r0 = rindex
tmp0 = tl.load(in_ptr0 + (r0), None)
tmp4 = tl.load(in_ptr1 + (r0), None)
tmp12 = tl.load(in_ptr0 + (64 + r0), None)
tmp16 = tl.load(in_ptr1 + (64 + r0), None)
tmp24 = tl.load(in_ptr0 + (128 + r0), None)
tmp28 = tl.load(in_ptr1 + (128 + r0), None)
tmp36 = tl.load(in_ptr0 + (192 + r0), None)
tmp40 = tl.load(in_ptr1 + (192 + r0), None)
tmp1 = tl.broadcast_to(tmp0, [XBLOCK, RBLOCK])
tmp3 = tl.sum(tmp1, 1)[:, None]
tmp5 = tl.broadcast_to(tmp4, [XBLOCK, RBLOCK])
tmp7 = tl.sum(tmp5, 1)[:, None]
tmp8 = tmp0 * tmp4
tmp9 = tl.broadcast_to(tmp8, [XBLOCK, RBLOCK])
tmp11 = tl.sum(tmp9, 1)[:, None]
tmp13 = tl.broadcast_to(tmp12, [XBLOCK, RBLOCK])
tmp15 = tl.sum(tmp13, 1)[:, None]
tmp17 = tl.broadcast_to(tmp16, [XBLOCK, RBLOCK])
tmp19 = tl.sum(tmp17, 1)[:, None]
tmp20 = tmp12 * tmp16
tmp21 = tl.broadcast_to(tmp20, [XBLOCK, RBLOCK])
tmp23 = tl.sum(tmp21, 1)[:, None]
tmp25 = tl.broadcast_to(tmp24, [XBLOCK, RBLOCK])
tmp27 = tl.sum(tmp25, 1)[:, None]
tmp29 = tl.broadcast_to(tmp28, [XBLOCK, RBLOCK])
tmp31 = tl.sum(tmp29, 1)[:, None]
tmp32 = tmp24 * tmp28
tmp33 = tl.broadcast_to(tmp32, [XBLOCK, RBLOCK])
tmp35 = tl.sum(tmp33, 1)[:, None]
tmp37 = tl.broadcast_to(tmp36, [XBLOCK, RBLOCK])
tmp39 = tl.sum(tmp37, 1)[:, None]
tmp41 = tl.broadcast_to(tmp40, [XBLOCK, RBLOCK])
tmp43 = tl.sum(tmp41, 1)[:, None]
tmp44 = tmp36 * tmp40
tmp45 = tl.broadcast_to(tmp44, [XBLOCK, RBLOCK])
tmp47 = tl.sum(tmp45, 1)[:, None]
tmp48 = tmp3 + tmp7
tmp49 = tmp48 - tmp11
tmp50 = tmp11 / tmp49
tmp51 = 1.0
tmp52 = tmp51 - tmp50
tmp53 = 0.0
tmp54 = tmp52 + tmp53
tmp55 = tmp15 + tmp19
tmp56 = tmp55 - tmp23
tmp57 = tmp23 / tmp56
tmp58 = tmp51 - tmp57
tmp59 = tmp54 + tmp58
tmp60 = tmp27 + tmp31
tmp61 = tmp60 - tmp35
tmp62 = tmp35 / tmp61
tmp63 = tmp51 - tmp62
tmp64 = tmp59 + tmp63
tmp65 = tmp39 + tmp43
tmp66 = tmp65 - tmp47
tmp67 = tmp47 / tmp66
tmp68 = tmp51 - tmp67
tmp69 = tmp64 + tmp68
tmp70 = 0.25
tmp71 = tmp69 * tmp70
tl.debug_barrier()
tl.store(in_out_ptr0 + (tl.full([XBLOCK, 1], 0, tl.int32)), tmp71, None)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, arg1_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg1_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((), (), torch.float32)
buf12 = buf0; del buf0 # reuse
buf13 = buf12; del buf12 # reuse
# Topologically Sorted Source Nodes: [sum_2, sum_3, add, mul, Iand1, Ior1, IoU1, sub_1, IoU, sum_5, sum_6, add_2, mul_1, Iand1_1, Ior1_1, IoU1_1, sub_3, IoU_1, sum_8, sum_9, add_4, mul_2, Iand1_2, Ior1_2, IoU1_2, sub_5, IoU_2, sum_11, sum_12, add_6, mul_3, Iand1_3, Ior1_3, IoU1_3, sub_7, IoU_3, truediv_4], Original ATen: [aten.sum, aten.add, aten.mul, aten.sub, aten.div, aten.rsub]
stream0 = get_raw_stream(0)
triton_per_fused_add_div_mul_rsub_sub_sum_0.run(buf13, arg1_1, arg0_1, 1, 64, grid=grid(1), stream=stream0)
del arg0_1
del arg1_1
return (buf13, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
arg1_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([arg0_1, arg1_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.multiprocessing
def _iou(pred, target, size_average=True):
b = pred.shape[0]
IoU = 0.0
for i in range(0, b):
Iand1 = torch.sum(target[i, :, :, :] * pred[i, :, :, :])
Ior1 = torch.sum(target[i, :, :, :]) + torch.sum(pred[i, :, :, :]
) - Iand1
IoU1 = Iand1 / Ior1
IoU = IoU + (1 - IoU1)
return IoU / b
class IOU(torch.nn.Module):
def __init__(self, size_average=True):
super(IOU, self).__init__()
self.size_average = size_average
def forward(self, pred, target):
return _iou(pred, target, self.size_average)
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {}]
| import torch
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
import torch.multiprocessing
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_per_fused_add_div_mul_rsub_sub_sum_0(in_out_ptr0, in_ptr0,
in_ptr1, xnumel, rnumel, XBLOCK: tl.constexpr):
RBLOCK: tl.constexpr = 64
xoffset = tl.program_id(0) * XBLOCK
xoffset + tl.arange(0, XBLOCK)[:, None]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[None, :]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
r0 = rindex
tmp0 = tl.load(in_ptr0 + r0, None)
tmp4 = tl.load(in_ptr1 + r0, None)
tmp12 = tl.load(in_ptr0 + (64 + r0), None)
tmp16 = tl.load(in_ptr1 + (64 + r0), None)
tmp24 = tl.load(in_ptr0 + (128 + r0), None)
tmp28 = tl.load(in_ptr1 + (128 + r0), None)
tmp36 = tl.load(in_ptr0 + (192 + r0), None)
tmp40 = tl.load(in_ptr1 + (192 + r0), None)
tmp1 = tl.broadcast_to(tmp0, [XBLOCK, RBLOCK])
tmp3 = tl.sum(tmp1, 1)[:, None]
tmp5 = tl.broadcast_to(tmp4, [XBLOCK, RBLOCK])
tmp7 = tl.sum(tmp5, 1)[:, None]
tmp8 = tmp0 * tmp4
tmp9 = tl.broadcast_to(tmp8, [XBLOCK, RBLOCK])
tmp11 = tl.sum(tmp9, 1)[:, None]
tmp13 = tl.broadcast_to(tmp12, [XBLOCK, RBLOCK])
tmp15 = tl.sum(tmp13, 1)[:, None]
tmp17 = tl.broadcast_to(tmp16, [XBLOCK, RBLOCK])
tmp19 = tl.sum(tmp17, 1)[:, None]
tmp20 = tmp12 * tmp16
tmp21 = tl.broadcast_to(tmp20, [XBLOCK, RBLOCK])
tmp23 = tl.sum(tmp21, 1)[:, None]
tmp25 = tl.broadcast_to(tmp24, [XBLOCK, RBLOCK])
tmp27 = tl.sum(tmp25, 1)[:, None]
tmp29 = tl.broadcast_to(tmp28, [XBLOCK, RBLOCK])
tmp31 = tl.sum(tmp29, 1)[:, None]
tmp32 = tmp24 * tmp28
tmp33 = tl.broadcast_to(tmp32, [XBLOCK, RBLOCK])
tmp35 = tl.sum(tmp33, 1)[:, None]
tmp37 = tl.broadcast_to(tmp36, [XBLOCK, RBLOCK])
tmp39 = tl.sum(tmp37, 1)[:, None]
tmp41 = tl.broadcast_to(tmp40, [XBLOCK, RBLOCK])
tmp43 = tl.sum(tmp41, 1)[:, None]
tmp44 = tmp36 * tmp40
tmp45 = tl.broadcast_to(tmp44, [XBLOCK, RBLOCK])
tmp47 = tl.sum(tmp45, 1)[:, None]
tmp48 = tmp3 + tmp7
tmp49 = tmp48 - tmp11
tmp50 = tmp11 / tmp49
tmp51 = 1.0
tmp52 = tmp51 - tmp50
tmp53 = 0.0
tmp54 = tmp52 + tmp53
tmp55 = tmp15 + tmp19
tmp56 = tmp55 - tmp23
tmp57 = tmp23 / tmp56
tmp58 = tmp51 - tmp57
tmp59 = tmp54 + tmp58
tmp60 = tmp27 + tmp31
tmp61 = tmp60 - tmp35
tmp62 = tmp35 / tmp61
tmp63 = tmp51 - tmp62
tmp64 = tmp59 + tmp63
tmp65 = tmp39 + tmp43
tmp66 = tmp65 - tmp47
tmp67 = tmp47 / tmp66
tmp68 = tmp51 - tmp67
tmp69 = tmp64 + tmp68
tmp70 = 0.25
tmp71 = tmp69 * tmp70
tl.debug_barrier()
tl.store(in_out_ptr0 + tl.full([XBLOCK, 1], 0, tl.int32), tmp71, None)
def call(args):
arg0_1, arg1_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg1_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((), (), torch.float32)
buf12 = buf0
del buf0
buf13 = buf12
del buf12
get_raw_stream(0)
triton_per_fused_add_div_mul_rsub_sub_sum_0[grid(1)](buf13, arg1_1,
arg0_1, 1, 64, XBLOCK=1, num_warps=2, num_stages=1)
del arg0_1
del arg1_1
return buf13,
def _iou(pred, target, size_average=True):
b = pred.shape[0]
IoU = 0.0
for i in range(0, b):
Iand1 = torch.sum(target[i, :, :, :] * pred[i, :, :, :])
Ior1 = torch.sum(target[i, :, :, :]) + torch.sum(pred[i, :, :, :]
) - Iand1
IoU1 = Iand1 / Ior1
IoU = IoU + (1 - IoU1)
return IoU / b
class IOUNew(torch.nn.Module):
def __init__(self, size_average=True):
super(IOUNew, self).__init__()
self.size_average = size_average
def forward(self, input_0, input_1):
arg0_1 = input_0
arg1_1 = input_1
output = call([arg0_1, arg1_1])
return output[0]
| intchous/SmartText | IOU | false | 6,886 | [
"MIT"
] | 1 | 81abb84ce135a3859c32257d861c9e87b51f8c3f | https://github.com/intchous/SmartText/tree/81abb84ce135a3859c32257d861c9e87b51f8c3f | import torch
import torch.multiprocessing
def _iou(pred, target, size_average=True):
b = pred.shape[0]
IoU = 0.0
for i in range(0, b):
Iand1 = torch.sum(target[i, :, :, :] * pred[i, :, :, :])
Ior1 = torch.sum(target[i, :, :, :]) + torch.sum(pred[i, :, :, :]
) - Iand1
IoU1 = Iand1 / Ior1
IoU = IoU + (1 - IoU1)
return IoU / b
class Model(torch.nn.Module):
def __init__(self, size_average=True):
super().__init__()
self.size_average = size_average
def forward(self, pred, target):
return _iou(pred, target, self.size_average)
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return []
|
CecaModule | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/ov/covbryzjnff2kb26c5gkcqbvct6kdwzanlx3iu6ee24itsit76o3.py
# Topologically Sorted Source Nodes: [mean], Original ATen: [aten.mean]
# Source node to ATen node mapping:
# mean => mean
# Graph fragment:
# %mean : [num_users=1] = call_function[target=torch.ops.aten.mean.dim](args = (%primals_1, [2, 3]), kwargs = {})
triton_per_fused_mean_0 = async_compile.triton('triton_per_fused_mean_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[16, 16],
reduction_hint=ReductionHint.INNER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused_mean_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 1, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused_mean_0(in_ptr0, out_ptr0, xnumel, rnumel, XBLOCK : tl.constexpr):
xnumel = 16
rnumel = 16
RBLOCK: tl.constexpr = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
roffset = 0
rmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
r1 = rindex
x0 = xindex
tmp0 = tl.load(in_ptr0 + (r1 + (16*x0)), xmask, other=0.0)
tmp1 = tl.broadcast_to(tmp0, [XBLOCK, RBLOCK])
tmp3 = tl.where(xmask, tmp1, 0)
tmp4 = tl.sum(tmp3, 1)[:, None]
tl.store(out_ptr0 + (x0), tmp4, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/5v/c5vy2mpdsk3ol4tdn7eyu5zzhzoueyp2ctrby2n6qikrbyckgsa7.py
# Topologically Sorted Source Nodes: [y_1], Original ATen: [aten.copy]
# Source node to ATen node mapping:
# y_1 => copy
# Graph fragment:
# %copy : [num_users=1] = call_function[target=torch.ops.aten.copy.default](args = (%slice_1, %slice_2), kwargs = {})
# %slice_scatter_default : [num_users=2] = call_function[target=torch.ops.aten.slice_scatter.default](args = (%empty, %copy, 2, 1, 5), kwargs = {})
# %slice_scatter_default_1 : [num_users=2] = call_function[target=torch.ops.aten.slice_scatter.default](args = (%slice_scatter_default, %slice_7, 2, 0, 1), kwargs = {})
# %slice_scatter_default_2 : [num_users=2] = call_function[target=torch.ops.aten.slice_scatter.default](args = (%slice_scatter_default_1, %slice_12, 2, 5, 6), kwargs = {})
triton_poi_fused_copy_1 = async_compile.triton('triton_poi_fused_copy_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[32],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_copy_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 4, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_copy_1(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 24
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 6
x1 = (xindex // 6)
x2 = xindex
tmp0 = x0
tmp1 = tl.full([1], 5, tl.int64)
tmp2 = tmp0 >= tmp1
tmp3 = (-4) + x0
tmp4 = tl.full([1], 1, tl.int64)
tmp5 = tmp3 < tmp4
tmp6 = tmp5 & tmp2
tmp7 = tmp0 >= tmp4
tmp8 = tmp0 < tmp1
tmp9 = tmp7 & tmp8
tmp10 = tmp9 & tmp6
tmp11 = tl.load(in_ptr0 + ((-1) + x0 + (4*x1)), tmp10 & xmask, other=0.0)
tmp12 = 16.0
tmp13 = tmp11 / tmp12
tmp14 = tl.full(tmp13.shape, 0.0, tmp13.dtype)
tmp15 = tl.where(tmp10, tmp13, tmp14)
tmp16 = float("nan")
tmp17 = tl.where(tmp9, tmp15, tmp16)
tmp18 = tl.full(tmp17.shape, 0.0, tmp17.dtype)
tmp19 = tl.where(tmp6, tmp17, tmp18)
tmp20 = tmp3 >= tmp4
tmp21 = tmp3 < tmp1
tmp22 = tmp20 & tmp21
tmp23 = tmp22 & tmp2
tmp24 = tl.load(in_ptr0 + ((-5) + x0 + (4*x1)), tmp23 & xmask, other=0.0)
tmp25 = tmp24 / tmp12
tmp26 = tl.full(tmp25.shape, 0.0, tmp25.dtype)
tmp27 = tl.where(tmp23, tmp25, tmp26)
tmp28 = tl.where(tmp22, tmp27, tmp16)
tmp29 = tl.where(tmp5, tmp19, tmp28)
tmp30 = tl.full(tmp29.shape, 0.0, tmp29.dtype)
tmp31 = tl.where(tmp2, tmp29, tmp30)
tmp32 = tmp0 < tmp4
tmp33 = 4 + x0
tmp34 = tmp33 >= tmp4
tmp35 = tmp33 < tmp1
tmp36 = tmp34 & tmp35
tmp37 = tmp36 & tmp32
tmp38 = tl.load(in_ptr0 + (3 + x0 + (4*x1)), tmp37 & xmask, other=0.0)
tmp39 = tmp38 / tmp12
tmp40 = tl.full(tmp39.shape, 0.0, tmp39.dtype)
tmp41 = tl.where(tmp37, tmp39, tmp40)
tmp42 = tl.where(tmp36, tmp41, tmp16)
tmp43 = tl.full(tmp42.shape, 0.0, tmp42.dtype)
tmp44 = tl.where(tmp32, tmp42, tmp43)
tmp45 = tl.load(in_ptr0 + ((-1) + x0 + (4*x1)), tmp9 & xmask, other=0.0)
tmp46 = tmp45 / tmp12
tmp47 = tl.full(tmp46.shape, 0.0, tmp46.dtype)
tmp48 = tl.where(tmp9, tmp46, tmp47)
tmp49 = tl.where(tmp9, tmp48, tmp16)
tmp50 = tl.where(tmp32, tmp44, tmp49)
tmp51 = tl.where(tmp2, tmp31, tmp50)
tl.store(out_ptr0 + (x2), tmp51, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/xv/cxvgsfj3x2o5ls6evsy4rhywutbtjkwezlavric3plphgvn75mea.py
# Topologically Sorted Source Nodes: [mul], Original ATen: [aten.mul]
# Source node to ATen node mapping:
# mul => mul
# Graph fragment:
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%primals_1, %expand), kwargs = {})
triton_poi_fused_mul_2 = async_compile.triton('triton_poi_fused_mul_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_mul_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_mul_2(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 16)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr1 + (x1), xmask, eviction_policy='evict_last')
tmp2 = tl.sigmoid(tmp1)
tmp3 = tmp0 * tmp2
tl.store(out_ptr0 + (x2), tmp3, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (1, 1, 3), (3, 3, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [mean], Original ATen: [aten.mean]
stream0 = get_raw_stream(0)
triton_per_fused_mean_0.run(primals_1, buf0, 16, 16, grid=grid(16), stream=stream0)
buf2 = empty_strided_cuda((4, 1, 6), (6, 6, 1), torch.float32)
# Topologically Sorted Source Nodes: [y_1], Original ATen: [aten.copy]
triton_poi_fused_copy_1.run(buf0, buf2, 24, grid=grid(24), stream=stream0)
del buf0
# Topologically Sorted Source Nodes: [y_2], Original ATen: [aten.convolution]
buf3 = extern_kernels.convolution(buf2, primals_2, stride=(1,), padding=(0,), dilation=(1,), transposed=False, output_padding=(0,), groups=1, bias=None)
assert_size_stride(buf3, (4, 1, 4), (4, 4, 1))
buf4 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [mul], Original ATen: [aten.mul]
triton_poi_fused_mul_2.run(primals_1, buf3, buf4, 256, grid=grid(256), stream=stream0)
return (buf4, primals_1, primals_2, buf2, buf3, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((1, 1, 3), (3, 3, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import math
import torch
import torch.nn.functional as F
import torch.nn as nn
class CecaModule(nn.Module):
"""Constructs a circular ECA module.
ECA module where the conv uses circular padding rather than zero padding.
Unlike the spatial dimension, the channels do not have inherent ordering nor
locality. Although this module in essence, applies such an assumption, it is unnecessary
to limit the channels on either "edge" from being circularly adapted to each other.
This will fundamentally increase connectivity and possibly increase performance metrics
(accuracy, robustness), without significantly impacting resource metrics
(parameter size, throughput,latency, etc)
Args:
channels: Number of channels of the input feature map for use in adaptive kernel sizes
for actual calculations according to channel.
gamma, beta: when channel is given parameters of mapping function
refer to original paper https://arxiv.org/pdf/1910.03151.pdf
(default=None. if channel size not given, use k_size given for kernel size.)
kernel_size: Adaptive selection of kernel size (default=3)
"""
def __init__(self, channels=None, kernel_size=3, gamma=2, beta=1):
super(CecaModule, self).__init__()
assert kernel_size % 2 == 1
if channels is not None:
t = int(abs(math.log(channels, 2) + beta) / gamma)
kernel_size = max(t if t % 2 else t + 1, 3)
self.conv = nn.Conv1d(1, 1, kernel_size=kernel_size, padding=0,
bias=False)
self.padding = (kernel_size - 1) // 2
def forward(self, x):
y = x.mean((2, 3)).view(x.shape[0], 1, -1)
y = F.pad(y, (self.padding, self.padding), mode='circular')
y = self.conv(y)
y = y.view(x.shape[0], -1, 1, 1).sigmoid()
return x * y.expand_as(x)
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
import math
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_per_fused_mean_0(in_ptr0, out_ptr0, xnumel, rnumel, XBLOCK: tl.
constexpr):
xnumel = 16
RBLOCK: tl.constexpr = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
r1 = rindex
x0 = xindex
tmp0 = tl.load(in_ptr0 + (r1 + 16 * x0), xmask, other=0.0)
tmp1 = tl.broadcast_to(tmp0, [XBLOCK, RBLOCK])
tmp3 = tl.where(xmask, tmp1, 0)
tmp4 = tl.sum(tmp3, 1)[:, None]
tl.store(out_ptr0 + x0, tmp4, xmask)
@triton.jit
def triton_poi_fused_copy_1(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 24
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 6
x1 = xindex // 6
x2 = xindex
tmp0 = x0
tmp1 = tl.full([1], 5, tl.int64)
tmp2 = tmp0 >= tmp1
tmp3 = -4 + x0
tmp4 = tl.full([1], 1, tl.int64)
tmp5 = tmp3 < tmp4
tmp6 = tmp5 & tmp2
tmp7 = tmp0 >= tmp4
tmp8 = tmp0 < tmp1
tmp9 = tmp7 & tmp8
tmp10 = tmp9 & tmp6
tmp11 = tl.load(in_ptr0 + (-1 + x0 + 4 * x1), tmp10 & xmask, other=0.0)
tmp12 = 16.0
tmp13 = tmp11 / tmp12
tmp14 = tl.full(tmp13.shape, 0.0, tmp13.dtype)
tmp15 = tl.where(tmp10, tmp13, tmp14)
tmp16 = float('nan')
tmp17 = tl.where(tmp9, tmp15, tmp16)
tmp18 = tl.full(tmp17.shape, 0.0, tmp17.dtype)
tmp19 = tl.where(tmp6, tmp17, tmp18)
tmp20 = tmp3 >= tmp4
tmp21 = tmp3 < tmp1
tmp22 = tmp20 & tmp21
tmp23 = tmp22 & tmp2
tmp24 = tl.load(in_ptr0 + (-5 + x0 + 4 * x1), tmp23 & xmask, other=0.0)
tmp25 = tmp24 / tmp12
tmp26 = tl.full(tmp25.shape, 0.0, tmp25.dtype)
tmp27 = tl.where(tmp23, tmp25, tmp26)
tmp28 = tl.where(tmp22, tmp27, tmp16)
tmp29 = tl.where(tmp5, tmp19, tmp28)
tmp30 = tl.full(tmp29.shape, 0.0, tmp29.dtype)
tmp31 = tl.where(tmp2, tmp29, tmp30)
tmp32 = tmp0 < tmp4
tmp33 = 4 + x0
tmp34 = tmp33 >= tmp4
tmp35 = tmp33 < tmp1
tmp36 = tmp34 & tmp35
tmp37 = tmp36 & tmp32
tmp38 = tl.load(in_ptr0 + (3 + x0 + 4 * x1), tmp37 & xmask, other=0.0)
tmp39 = tmp38 / tmp12
tmp40 = tl.full(tmp39.shape, 0.0, tmp39.dtype)
tmp41 = tl.where(tmp37, tmp39, tmp40)
tmp42 = tl.where(tmp36, tmp41, tmp16)
tmp43 = tl.full(tmp42.shape, 0.0, tmp42.dtype)
tmp44 = tl.where(tmp32, tmp42, tmp43)
tmp45 = tl.load(in_ptr0 + (-1 + x0 + 4 * x1), tmp9 & xmask, other=0.0)
tmp46 = tmp45 / tmp12
tmp47 = tl.full(tmp46.shape, 0.0, tmp46.dtype)
tmp48 = tl.where(tmp9, tmp46, tmp47)
tmp49 = tl.where(tmp9, tmp48, tmp16)
tmp50 = tl.where(tmp32, tmp44, tmp49)
tmp51 = tl.where(tmp2, tmp31, tmp50)
tl.store(out_ptr0 + x2, tmp51, xmask)
@triton.jit
def triton_poi_fused_mul_2(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK: tl.
constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 16
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr1 + x1, xmask, eviction_policy='evict_last')
tmp2 = tl.sigmoid(tmp1)
tmp3 = tmp0 * tmp2
tl.store(out_ptr0 + x2, tmp3, xmask)
def call(args):
primals_1, primals_2 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (1, 1, 3), (3, 3, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
get_raw_stream(0)
triton_per_fused_mean_0[grid(16)](primals_1, buf0, 16, 16, XBLOCK=1,
num_warps=2, num_stages=1)
buf2 = empty_strided_cuda((4, 1, 6), (6, 6, 1), torch.float32)
triton_poi_fused_copy_1[grid(24)](buf0, buf2, 24, XBLOCK=32,
num_warps=1, num_stages=1)
del buf0
buf3 = extern_kernels.convolution(buf2, primals_2, stride=(1,),
padding=(0,), dilation=(1,), transposed=False, output_padding=(
0,), groups=1, bias=None)
assert_size_stride(buf3, (4, 1, 4), (4, 4, 1))
buf4 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
triton_poi_fused_mul_2[grid(256)](primals_1, buf3, buf4, 256,
XBLOCK=256, num_warps=4, num_stages=1)
return buf4, primals_1, primals_2, buf2, buf3
class CecaModuleNew(nn.Module):
"""Constructs a circular ECA module.
ECA module where the conv uses circular padding rather than zero padding.
Unlike the spatial dimension, the channels do not have inherent ordering nor
locality. Although this module in essence, applies such an assumption, it is unnecessary
to limit the channels on either "edge" from being circularly adapted to each other.
This will fundamentally increase connectivity and possibly increase performance metrics
(accuracy, robustness), without significantly impacting resource metrics
(parameter size, throughput,latency, etc)
Args:
channels: Number of channels of the input feature map for use in adaptive kernel sizes
for actual calculations according to channel.
gamma, beta: when channel is given parameters of mapping function
refer to original paper https://arxiv.org/pdf/1910.03151.pdf
(default=None. if channel size not given, use k_size given for kernel size.)
kernel_size: Adaptive selection of kernel size (default=3)
"""
def __init__(self, channels=None, kernel_size=3, gamma=2, beta=1):
super(CecaModuleNew, self).__init__()
assert kernel_size % 2 == 1
if channels is not None:
t = int(abs(math.log(channels, 2) + beta) / gamma)
kernel_size = max(t if t % 2 else t + 1, 3)
self.conv = nn.Conv1d(1, 1, kernel_size=kernel_size, padding=0,
bias=False)
self.padding = (kernel_size - 1) // 2
def forward(self, input_0):
primals_2 = self.conv.weight
primals_1 = input_0
output = call([primals_1, primals_2])
return output[0]
| infomon/meta_nas | CecaModule | false | 6,887 | [
"Apache-2.0"
] | 1 | b81b7de86d26ae1ec0d6646b4277f3c918e5e35d | https://github.com/infomon/meta_nas/tree/b81b7de86d26ae1ec0d6646b4277f3c918e5e35d | import math
import torch
import torch.nn.functional as F
import torch.nn as nn
class Model(nn.Module):
"""Constructs a circular ECA module.
ECA module where the conv uses circular padding rather than zero padding.
Unlike the spatial dimension, the channels do not have inherent ordering nor
locality. Although this module in essence, applies such an assumption, it is unnecessary
to limit the channels on either "edge" from being circularly adapted to each other.
This will fundamentally increase connectivity and possibly increase performance metrics
(accuracy, robustness), without significantly impacting resource metrics
(parameter size, throughput,latency, etc)
Args:
channels: Number of channels of the input feature map for use in adaptive kernel sizes
for actual calculations according to channel.
gamma, beta: when channel is given parameters of mapping function
refer to original paper https://arxiv.org/pdf/1910.03151.pdf
(default=None. if channel size not given, use k_size given for kernel size.)
kernel_size: Adaptive selection of kernel size (default=3)
"""
def __init__(self, channels=None, kernel_size=3, gamma=2, beta=1):
super().__init__()
assert kernel_size % 2 == 1
if channels is not None:
t = int(abs(math.log(channels, 2) + beta) / gamma)
kernel_size = max(t if t % 2 else t + 1, 3)
self.conv = nn.Conv1d(1, 1, kernel_size=kernel_size, padding=0,
bias=False)
self.padding = (kernel_size - 1) // 2
def forward(self, x):
y = x.mean((2, 3)).view(x.shape[0], 1, -1)
y = F.pad(y, (self.padding, self.padding), mode='circular')
y = self.conv(y)
y = y.view(x.shape[0], -1, 1, 1).sigmoid()
return x * y.expand_as(x)
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return []
|
Binarizer | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/js/cjsc3zpl3hrqyuqwese3sr3hjo7bt5zaivgps44t4f7tcgplnrpv.py
# Topologically Sorted Source Nodes: [gt, float_1], Original ATen: [aten.gt, aten._to_copy]
# Source node to ATen node mapping:
# float_1 => convert_element_type
# gt => gt
# Graph fragment:
# %gt : [num_users=1] = call_function[target=torch.ops.aten.gt.Tensor](args = (%arg1_1, %arg0_1), kwargs = {})
# %convert_element_type : [num_users=1] = call_function[target=torch.ops.prims.convert_element_type.default](args = (%gt, torch.float32), kwargs = {})
triton_poi_fused__to_copy_gt_0 = async_compile.triton('triton_poi_fused__to_copy_gt_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__to_copy_gt_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__to_copy_gt_0(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (x0), xmask)
tmp1 = tl.load(in_ptr1 + (0))
tmp2 = tl.broadcast_to(tmp1, [XBLOCK])
tmp3 = tmp0 > tmp2
tmp4 = tmp3.to(tl.float32)
tl.store(out_ptr0 + (x0), tmp4, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, arg1_1 = args
args.clear()
assert_size_stride(arg0_1, (1, ), (1, ))
assert_size_stride(arg1_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [gt, float_1], Original ATen: [aten.gt, aten._to_copy]
stream0 = get_raw_stream(0)
triton_poi_fused__to_copy_gt_0.run(arg1_1, arg0_1, buf0, 256, grid=grid(256), stream=stream0)
del arg0_1
del arg1_1
return (buf0, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((1, ), (1, ), device='cuda:0', dtype=torch.float32)
arg1_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([arg0_1, arg1_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
from abc import ABC
from sklearn.preprocessing import Binarizer
class BaseOperator(ABC):
"""
Abstract class defining the basic structure for operator implementations in Hummingbird.
"""
def __init__(self, regression=False, classification=False, transformer=
False, anomaly_detection=False, **kwargs):
"""
Args:
regression: Whether operator is a regression model.
classification: Whether the operator is a classification model.
transformer: Whether the operator is a feature transformer.
anomaly_detection: Whether the operator is an anomaly detection model.
kwargs: Other keyword arguments.
"""
super().__init__()
self.regression = regression
self.classification = classification
self.transformer = transformer
self.anomaly_detection = anomaly_detection
class Binarizer(BaseOperator, torch.nn.Module):
"""
Class implementing Binarizer operators in PyTorch.
"""
def __init__(self, threshold, device):
super(Binarizer, self).__init__()
self.transformer = True
self.threshold = torch.nn.Parameter(torch.FloatTensor([threshold]),
requires_grad=False)
def forward(self, x):
return torch.gt(x, self.threshold).float()
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'threshold': 4, 'device': 0}]
| import torch
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from abc import ABC
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_poi_fused__to_copy_gt_0(in_ptr0, in_ptr1, out_ptr0, xnumel,
XBLOCK: tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + x0, xmask)
tmp1 = tl.load(in_ptr1 + 0)
tmp2 = tl.broadcast_to(tmp1, [XBLOCK])
tmp3 = tmp0 > tmp2
tmp4 = tmp3.to(tl.float32)
tl.store(out_ptr0 + x0, tmp4, xmask)
def call(args):
arg0_1, arg1_1 = args
args.clear()
assert_size_stride(arg0_1, (1,), (1,))
assert_size_stride(arg1_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused__to_copy_gt_0[grid(256)](arg1_1, arg0_1, buf0, 256,
XBLOCK=256, num_warps=4, num_stages=1)
del arg0_1
del arg1_1
return buf0,
class BaseOperator(ABC):
"""
Abstract class defining the basic structure for operator implementations in Hummingbird.
"""
def __init__(self, regression=False, classification=False, transformer=
False, anomaly_detection=False, **kwargs):
"""
Args:
regression: Whether operator is a regression model.
classification: Whether the operator is a classification model.
transformer: Whether the operator is a feature transformer.
anomaly_detection: Whether the operator is an anomaly detection model.
kwargs: Other keyword arguments.
"""
super().__init__()
self.regression = regression
self.classification = classification
self.transformer = transformer
self.anomaly_detection = anomaly_detection
class BinarizerNew(BaseOperator, torch.nn.Module):
"""
Class implementing Binarizer operators in PyTorch.
"""
def __init__(self, threshold, device):
super(BinarizerNew, self).__init__()
self.transformer = True
self.threshold = torch.nn.Parameter(torch.FloatTensor([threshold]),
requires_grad=False)
def forward(self, input_0):
arg0_1 = self.threshold
arg1_1 = input_0
output = call([arg0_1, arg1_1])
return output[0]
| hannahaih/hummingbird | Binarizer | false | 6,888 | [
"MIT"
] | 1 | b8ec670b3c90ec7e87d3ae4a2b268075bd5eae65 | https://github.com/hannahaih/hummingbird/tree/b8ec670b3c90ec7e87d3ae4a2b268075bd5eae65 | import torch
from abc import ABC
from sklearn.preprocessing import Binarizer
class BaseOperator(ABC):
"""
Abstract class defining the basic structure for operator implementations in Hummingbird.
"""
def __init__(self, regression=False, classification=False, transformer=
False, anomaly_detection=False, **kwargs):
"""
Args:
regression: Whether operator is a regression model.
classification: Whether the operator is a classification model.
transformer: Whether the operator is a feature transformer.
anomaly_detection: Whether the operator is an anomaly detection model.
kwargs: Other keyword arguments.
"""
super().__init__()
self.regression = regression
self.classification = classification
self.transformer = transformer
self.anomaly_detection = anomaly_detection
class Model(BaseOperator, torch.nn.Module):
"""
Class implementing Binarizer operators in PyTorch.
"""
def __init__(self, threshold, device):
super().__init__()
self.transformer = True
self.threshold = torch.nn.Parameter(torch.FloatTensor([threshold]),
requires_grad=False)
def forward(self, x):
return torch.gt(x, self.threshold).float()
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4, 0]
|
ActorNetwork | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/hp/chpdwpegv6lvistek2wqgimtufecqvfp6grp5rpblk5yjicjzqd2.py
# Topologically Sorted Source Nodes: [layer_norm], Original ATen: [aten.native_layer_norm]
# Source node to ATen node mapping:
# layer_norm => add, rsqrt, var_mean
# Graph fragment:
# %var_mean : [num_users=2] = call_function[target=torch.ops.aten.var_mean.correction](args = (%view_1, [3]), kwargs = {correction: 0, keepdim: True})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%getitem, 1e-05), kwargs = {})
# %rsqrt : [num_users=1] = call_function[target=torch.ops.aten.rsqrt.default](args = (%add,), kwargs = {})
triton_poi_fused_native_layer_norm_0 = async_compile.triton('triton_poi_fused_native_layer_norm_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_native_layer_norm_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 4, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_native_layer_norm_0(in_ptr0, out_ptr0, out_ptr1, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (4*x0), xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr0 + (1 + (4*x0)), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr0 + (2 + (4*x0)), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (3 + (4*x0)), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp4 = tmp2 + tmp3
tmp6 = tmp4 + tmp5
tmp7 = 4.0
tmp8 = tmp6 / tmp7
tmp9 = tmp0 - tmp8
tmp10 = tmp9 * tmp9
tmp11 = tmp1 - tmp8
tmp12 = tmp11 * tmp11
tmp13 = tmp10 + tmp12
tmp14 = tmp3 - tmp8
tmp15 = tmp14 * tmp14
tmp16 = tmp13 + tmp15
tmp17 = tmp5 - tmp8
tmp18 = tmp17 * tmp17
tmp19 = tmp16 + tmp18
tmp20 = tmp19 / tmp7
tmp21 = 1e-05
tmp22 = tmp20 + tmp21
tmp23 = libdevice.rsqrt(tmp22)
tl.store(out_ptr0 + (x0), tmp8, xmask)
tl.store(out_ptr1 + (x0), tmp23, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/6e/c6etrnrwnqlbwohg25wyqa2rkojloi6lu4ex24dati4k5ok6ywe4.py
# Topologically Sorted Source Nodes: [layer_norm, x], Original ATen: [aten.native_layer_norm, aten.relu, aten.threshold_backward]
# Source node to ATen node mapping:
# layer_norm => add, add_1, mul, mul_1, rsqrt, sub, var_mean
# x => relu
# Graph fragment:
# %var_mean : [num_users=2] = call_function[target=torch.ops.aten.var_mean.correction](args = (%view_1, [3]), kwargs = {correction: 0, keepdim: True})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%getitem, 1e-05), kwargs = {})
# %rsqrt : [num_users=1] = call_function[target=torch.ops.aten.rsqrt.default](args = (%add,), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%view_1, %getitem_1), kwargs = {})
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sub, %rsqrt), kwargs = {})
# %mul_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%mul, %primals_4), kwargs = {})
# %add_1 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul_1, %primals_5), kwargs = {})
# %relu : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%add_1,), kwargs = {})
# %le_1 : [num_users=1] = call_function[target=torch.ops.aten.le.Scalar](args = (%relu, 0), kwargs = {})
triton_poi_fused_native_layer_norm_relu_threshold_backward_1 = async_compile.triton('triton_poi_fused_native_layer_norm_relu_threshold_backward_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: '*fp32', 5: '*fp32', 6: '*i1', 7: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5, 6, 7), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_native_layer_norm_relu_threshold_backward_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_native_layer_norm_relu_threshold_backward_1(in_ptr0, in_ptr1, in_ptr2, in_ptr3, in_ptr4, out_ptr0, out_ptr1, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
x0 = xindex % 4
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr1 + (x1), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr2 + (x1), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr3 + (x0), xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr4 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 - tmp1
tmp4 = tmp2 * tmp3
tmp6 = tmp4 * tmp5
tmp8 = tmp6 + tmp7
tmp9 = tl.full([1], 0, tl.int32)
tmp10 = triton_helpers.maximum(tmp9, tmp8)
tmp11 = 0.0
tmp12 = tmp10 <= tmp11
tl.store(out_ptr0 + (x2), tmp10, xmask)
tl.store(out_ptr1 + (x2), tmp12, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/ns/cnszijuiz432ctw37rqktvk3syr2vugzeuatmva3neoizic6f3sq.py
# Topologically Sorted Source Nodes: [action], Original ATen: [aten.tanh]
# Source node to ATen node mapping:
# action => tanh
# Graph fragment:
# %tanh : [num_users=1] = call_function[target=torch.ops.aten.tanh.default](args = (%view_5,), kwargs = {})
triton_poi_fused_tanh_2 = async_compile.triton('triton_poi_fused_tanh_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_tanh_2', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_tanh_2(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = libdevice.tanh(tmp2)
tl.store(in_out_ptr0 + (x2), tmp3, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9, primals_10, primals_11 = args
args.clear()
assert_size_stride(primals_1, (4, 4), (4, 1))
assert_size_stride(primals_2, (4, ), (1, ))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_4, (4, ), (1, ))
assert_size_stride(primals_5, (4, ), (1, ))
assert_size_stride(primals_6, (4, 4), (4, 1))
assert_size_stride(primals_7, (4, ), (1, ))
assert_size_stride(primals_8, (4, ), (1, ))
assert_size_stride(primals_9, (4, ), (1, ))
assert_size_stride(primals_10, (4, 4), (4, 1))
assert_size_stride(primals_11, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [linear], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_2, reinterpret_tensor(primals_3, (64, 4), (4, 1), 0), reinterpret_tensor(primals_1, (4, 4), (1, 4), 0), alpha=1, beta=1, out=buf0)
del primals_1
del primals_2
buf1 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 64), torch.float32)
buf2 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 64), torch.float32)
# Topologically Sorted Source Nodes: [layer_norm], Original ATen: [aten.native_layer_norm]
stream0 = get_raw_stream(0)
triton_poi_fused_native_layer_norm_0.run(buf0, buf1, buf2, 64, grid=grid(64), stream=stream0)
buf3 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
buf11 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.bool)
# Topologically Sorted Source Nodes: [layer_norm, x], Original ATen: [aten.native_layer_norm, aten.relu, aten.threshold_backward]
triton_poi_fused_native_layer_norm_relu_threshold_backward_1.run(buf0, buf1, buf2, primals_4, primals_5, buf3, buf11, 256, grid=grid(256), stream=stream0)
del primals_5
buf4 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [linear_1], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_7, reinterpret_tensor(buf3, (64, 4), (4, 1), 0), reinterpret_tensor(primals_6, (4, 4), (1, 4), 0), alpha=1, beta=1, out=buf4)
del primals_7
buf5 = buf2; del buf2 # reuse
buf6 = buf1; del buf1 # reuse
# Topologically Sorted Source Nodes: [layer_norm_1], Original ATen: [aten.native_layer_norm]
triton_poi_fused_native_layer_norm_0.run(buf4, buf5, buf6, 64, grid=grid(64), stream=stream0)
buf7 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
buf10 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.bool)
# Topologically Sorted Source Nodes: [layer_norm_1, x_1], Original ATen: [aten.native_layer_norm, aten.relu, aten.threshold_backward]
triton_poi_fused_native_layer_norm_relu_threshold_backward_1.run(buf4, buf5, buf6, primals_8, primals_9, buf7, buf10, 256, grid=grid(256), stream=stream0)
del buf5
del buf6
del primals_9
buf8 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(buf7, (64, 4), (4, 1), 0), reinterpret_tensor(primals_10, (4, 4), (1, 4), 0), out=buf8)
buf9 = reinterpret_tensor(buf8, (4, 4, 4, 4), (64, 16, 4, 1), 0); del buf8 # reuse
# Topologically Sorted Source Nodes: [action], Original ATen: [aten.tanh]
triton_poi_fused_tanh_2.run(buf9, primals_11, 256, grid=grid(256), stream=stream0)
del primals_11
return (buf9, primals_4, primals_8, reinterpret_tensor(primals_3, (64, 4), (4, 1), 0), buf0, reinterpret_tensor(buf3, (64, 4), (4, 1), 0), buf4, reinterpret_tensor(buf7, (64, 4), (4, 1), 0), buf9, primals_10, buf10, primals_6, buf11, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_8 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_9 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_10 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_11 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9, primals_10, primals_11])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch as T
import torch.nn as nn
import torch.optim as optim
class ActorNetwork(nn.Module):
def __init__(self, alpha, state_dim, action_dim, fc1_dim, fc2_dim):
super(ActorNetwork, self).__init__()
self.fc1 = nn.Linear(state_dim, fc1_dim)
self.ln1 = nn.LayerNorm(fc1_dim)
self.fc2 = nn.Linear(fc1_dim, fc2_dim)
self.ln2 = nn.LayerNorm(fc2_dim)
self.action = nn.Linear(fc2_dim, action_dim)
self.optimizer = optim.Adam(self.parameters(), lr=alpha)
self
def forward(self, state):
x = T.relu(self.ln1(self.fc1(state)))
x = T.relu(self.ln2(self.fc2(x)))
action = T.tanh(self.action(x))
return action
def save_checkpoint(self, checkpoint_file):
T.save(self.state_dict(), checkpoint_file,
_use_new_zipfile_serialization=False)
def load_checkpoint(self, checkpoint_file):
self.load_state_dict(T.load(checkpoint_file))
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'alpha': 4, 'state_dim': 4, 'action_dim': 4, 'fc1_dim': 4,
'fc2_dim': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import libdevice
import torch as T
import torch.nn as nn
import torch.optim as optim
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_native_layer_norm_0(in_ptr0, out_ptr0, out_ptr1,
xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + 4 * x0, xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr0 + (1 + 4 * x0), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr0 + (2 + 4 * x0), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (3 + 4 * x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp4 = tmp2 + tmp3
tmp6 = tmp4 + tmp5
tmp7 = 4.0
tmp8 = tmp6 / tmp7
tmp9 = tmp0 - tmp8
tmp10 = tmp9 * tmp9
tmp11 = tmp1 - tmp8
tmp12 = tmp11 * tmp11
tmp13 = tmp10 + tmp12
tmp14 = tmp3 - tmp8
tmp15 = tmp14 * tmp14
tmp16 = tmp13 + tmp15
tmp17 = tmp5 - tmp8
tmp18 = tmp17 * tmp17
tmp19 = tmp16 + tmp18
tmp20 = tmp19 / tmp7
tmp21 = 1e-05
tmp22 = tmp20 + tmp21
tmp23 = libdevice.rsqrt(tmp22)
tl.store(out_ptr0 + x0, tmp8, xmask)
tl.store(out_ptr1 + x0, tmp23, xmask)
@triton.jit
def triton_poi_fused_native_layer_norm_relu_threshold_backward_1(in_ptr0,
in_ptr1, in_ptr2, in_ptr3, in_ptr4, out_ptr0, out_ptr1, xnumel, XBLOCK:
tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
x0 = xindex % 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr1 + x1, xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr2 + x1, xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr3 + x0, xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr4 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 - tmp1
tmp4 = tmp2 * tmp3
tmp6 = tmp4 * tmp5
tmp8 = tmp6 + tmp7
tmp9 = tl.full([1], 0, tl.int32)
tmp10 = triton_helpers.maximum(tmp9, tmp8)
tmp11 = 0.0
tmp12 = tmp10 <= tmp11
tl.store(out_ptr0 + x2, tmp10, xmask)
tl.store(out_ptr1 + x2, tmp12, xmask)
@triton.jit
def triton_poi_fused_tanh_2(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = libdevice.tanh(tmp2)
tl.store(in_out_ptr0 + x2, tmp3, xmask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7, primals_8, primals_9, primals_10, primals_11) = args
args.clear()
assert_size_stride(primals_1, (4, 4), (4, 1))
assert_size_stride(primals_2, (4,), (1,))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_4, (4,), (1,))
assert_size_stride(primals_5, (4,), (1,))
assert_size_stride(primals_6, (4, 4), (4, 1))
assert_size_stride(primals_7, (4,), (1,))
assert_size_stride(primals_8, (4,), (1,))
assert_size_stride(primals_9, (4,), (1,))
assert_size_stride(primals_10, (4, 4), (4, 1))
assert_size_stride(primals_11, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
extern_kernels.addmm(primals_2, reinterpret_tensor(primals_3, (64,
4), (4, 1), 0), reinterpret_tensor(primals_1, (4, 4), (1, 4), 0
), alpha=1, beta=1, out=buf0)
del primals_1
del primals_2
buf1 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 64), torch.float32)
buf2 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 64), torch.float32)
get_raw_stream(0)
triton_poi_fused_native_layer_norm_0[grid(64)](buf0, buf1, buf2, 64,
XBLOCK=64, num_warps=1, num_stages=1)
buf3 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
buf11 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.bool)
triton_poi_fused_native_layer_norm_relu_threshold_backward_1[grid(256)
](buf0, buf1, buf2, primals_4, primals_5, buf3, buf11, 256,
XBLOCK=128, num_warps=4, num_stages=1)
del primals_5
buf4 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
extern_kernels.addmm(primals_7, reinterpret_tensor(buf3, (64, 4), (
4, 1), 0), reinterpret_tensor(primals_6, (4, 4), (1, 4), 0),
alpha=1, beta=1, out=buf4)
del primals_7
buf5 = buf2
del buf2
buf6 = buf1
del buf1
triton_poi_fused_native_layer_norm_0[grid(64)](buf4, buf5, buf6, 64,
XBLOCK=64, num_warps=1, num_stages=1)
buf7 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
buf10 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.bool)
triton_poi_fused_native_layer_norm_relu_threshold_backward_1[grid(256)
](buf4, buf5, buf6, primals_8, primals_9, buf7, buf10, 256,
XBLOCK=128, num_warps=4, num_stages=1)
del buf5
del buf6
del primals_9
buf8 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(buf7, (64, 4), (4, 1), 0),
reinterpret_tensor(primals_10, (4, 4), (1, 4), 0), out=buf8)
buf9 = reinterpret_tensor(buf8, (4, 4, 4, 4), (64, 16, 4, 1), 0)
del buf8
triton_poi_fused_tanh_2[grid(256)](buf9, primals_11, 256, XBLOCK=
128, num_warps=4, num_stages=1)
del primals_11
return buf9, primals_4, primals_8, reinterpret_tensor(primals_3, (64, 4
), (4, 1), 0), buf0, reinterpret_tensor(buf3, (64, 4), (4, 1), 0
), buf4, reinterpret_tensor(buf7, (64, 4), (4, 1), 0
), buf9, primals_10, buf10, primals_6, buf11
class ActorNetworkNew(nn.Module):
def __init__(self, alpha, state_dim, action_dim, fc1_dim, fc2_dim):
super(ActorNetworkNew, self).__init__()
self.fc1 = nn.Linear(state_dim, fc1_dim)
self.ln1 = nn.LayerNorm(fc1_dim)
self.fc2 = nn.Linear(fc1_dim, fc2_dim)
self.ln2 = nn.LayerNorm(fc2_dim)
self.action = nn.Linear(fc2_dim, action_dim)
self.optimizer = optim.Adam(self.parameters(), lr=alpha)
self
def save_checkpoint(self, checkpoint_file):
T.save(self.state_dict(), checkpoint_file,
_use_new_zipfile_serialization=False)
def load_checkpoint(self, checkpoint_file):
self.load_state_dict(T.load(checkpoint_file))
def forward(self, input_0):
primals_1 = self.fc1.weight
primals_2 = self.fc1.bias
primals_4 = self.ln1.weight
primals_5 = self.ln1.bias
primals_6 = self.fc2.weight
primals_7 = self.fc2.bias
primals_8 = self.ln2.weight
primals_9 = self.ln2.bias
primals_10 = self.action.weight
primals_11 = self.action.bias
primals_3 = input_0
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7, primals_8, primals_9,
primals_10, primals_11])
return output[0]
| indigoLovee/TD3 | ActorNetwork | false | 6,889 | [
"MIT"
] | 1 | 0e86a40c27ec376b52e9f8e0e70db28e7411276b | https://github.com/indigoLovee/TD3/tree/0e86a40c27ec376b52e9f8e0e70db28e7411276b | import torch
import torch as T
import torch.nn as nn
import torch.optim as optim
class Model(nn.Module):
def __init__(self, alpha, state_dim, action_dim, fc1_dim, fc2_dim):
super().__init__()
self.fc1 = nn.Linear(state_dim, fc1_dim)
self.ln1 = nn.LayerNorm(fc1_dim)
self.fc2 = nn.Linear(fc1_dim, fc2_dim)
self.ln2 = nn.LayerNorm(fc2_dim)
self.action = nn.Linear(fc2_dim, action_dim)
self.optimizer = optim.Adam(self.parameters(), lr=alpha)
self
def forward(self, state):
x = T.relu(self.ln1(self.fc1(state)))
x = T.relu(self.ln2(self.fc2(x)))
action = T.tanh(self.action(x))
return action
def save_checkpoint(self, checkpoint_file):
T.save(self.state_dict(), checkpoint_file,
_use_new_zipfile_serialization=False)
def load_checkpoint(self, checkpoint_file):
self.load_state_dict(T.load(checkpoint_file))
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'alpha': 4, 'state_dim': 4, 'action_dim': 4, 'fc1_dim': 4,
'fc2_dim': 4}]
|
CriticNetwork | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/c4/cc4khg7fwbxxm2fufox7nnkf4gfybrmj5ir2tx3zuxfioc5b2dya.py
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.cat]
# Source node to ATen node mapping:
# x => cat
# Graph fragment:
# %cat : [num_users=1] = call_function[target=torch.ops.aten.cat.default](args = ([%primals_1, %primals_2], -1), kwargs = {})
triton_poi_fused_cat_0 = async_compile.triton('triton_poi_fused_cat_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[512],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_cat_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_cat_0(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 512
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 8
x1 = (xindex // 8)
x2 = xindex
tmp0 = x0
tmp1 = tl.full([1], 0, tl.int64)
tmp2 = tmp0 >= tmp1
tmp3 = tl.full([1], 4, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = tl.load(in_ptr0 + ((4*x1) + x0), tmp4 & xmask, eviction_policy='evict_last', other=0.0)
tmp6 = tmp0 >= tmp3
tmp7 = tl.full([1], 8, tl.int64)
tmp8 = tmp0 < tmp7
tmp9 = tl.load(in_ptr1 + ((4*x1) + ((-4) + x0)), tmp6 & xmask, eviction_policy='evict_last', other=0.0)
tmp10 = tl.where(tmp4, tmp5, tmp9)
tl.store(out_ptr0 + (x2), tmp10, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/6n/c6nwltytpo33ssumvxlcryrpvlql2hsjrmxl624j4dkkjxt5qgkm.py
# Topologically Sorted Source Nodes: [layer_norm], Original ATen: [aten.native_layer_norm]
# Source node to ATen node mapping:
# layer_norm => add, rsqrt, var_mean
# Graph fragment:
# %var_mean : [num_users=2] = call_function[target=torch.ops.aten.var_mean.correction](args = (%view_1, [3]), kwargs = {correction: 0, keepdim: True})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%getitem, 1e-05), kwargs = {})
# %rsqrt : [num_users=1] = call_function[target=torch.ops.aten.rsqrt.default](args = (%add,), kwargs = {})
triton_poi_fused_native_layer_norm_1 = async_compile.triton('triton_poi_fused_native_layer_norm_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_native_layer_norm_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 4, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_native_layer_norm_1(in_ptr0, out_ptr0, out_ptr1, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (4*x0), xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr0 + (1 + (4*x0)), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr0 + (2 + (4*x0)), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (3 + (4*x0)), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp4 = tmp2 + tmp3
tmp6 = tmp4 + tmp5
tmp7 = 4.0
tmp8 = tmp6 / tmp7
tmp9 = tmp0 - tmp8
tmp10 = tmp9 * tmp9
tmp11 = tmp1 - tmp8
tmp12 = tmp11 * tmp11
tmp13 = tmp10 + tmp12
tmp14 = tmp3 - tmp8
tmp15 = tmp14 * tmp14
tmp16 = tmp13 + tmp15
tmp17 = tmp5 - tmp8
tmp18 = tmp17 * tmp17
tmp19 = tmp16 + tmp18
tmp20 = tmp19 / tmp7
tmp21 = 1e-05
tmp22 = tmp20 + tmp21
tmp23 = libdevice.rsqrt(tmp22)
tl.store(out_ptr0 + (x0), tmp8, xmask)
tl.store(out_ptr1 + (x0), tmp23, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/hc/chcjzdcylbmtoa47cdtoahlodvthwl2d2u7dieocbyxknehyzic2.py
# Topologically Sorted Source Nodes: [layer_norm, x_1], Original ATen: [aten.native_layer_norm, aten.relu, aten.threshold_backward]
# Source node to ATen node mapping:
# layer_norm => add, add_1, mul, mul_1, rsqrt, sub, var_mean
# x_1 => relu
# Graph fragment:
# %var_mean : [num_users=2] = call_function[target=torch.ops.aten.var_mean.correction](args = (%view_1, [3]), kwargs = {correction: 0, keepdim: True})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%getitem, 1e-05), kwargs = {})
# %rsqrt : [num_users=1] = call_function[target=torch.ops.aten.rsqrt.default](args = (%add,), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%view_1, %getitem_1), kwargs = {})
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sub, %rsqrt), kwargs = {})
# %mul_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%mul, %primals_5), kwargs = {})
# %add_1 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul_1, %primals_6), kwargs = {})
# %relu : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%add_1,), kwargs = {})
# %le_1 : [num_users=1] = call_function[target=torch.ops.aten.le.Scalar](args = (%relu, 0), kwargs = {})
triton_poi_fused_native_layer_norm_relu_threshold_backward_2 = async_compile.triton('triton_poi_fused_native_layer_norm_relu_threshold_backward_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: '*fp32', 5: '*fp32', 6: '*i1', 7: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5, 6, 7), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_native_layer_norm_relu_threshold_backward_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_native_layer_norm_relu_threshold_backward_2(in_ptr0, in_ptr1, in_ptr2, in_ptr3, in_ptr4, out_ptr0, out_ptr1, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
x0 = xindex % 4
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr1 + (x1), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr2 + (x1), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr3 + (x0), xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr4 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 - tmp1
tmp4 = tmp2 * tmp3
tmp6 = tmp4 * tmp5
tmp8 = tmp6 + tmp7
tmp9 = tl.full([1], 0, tl.int32)
tmp10 = triton_helpers.maximum(tmp9, tmp8)
tmp11 = 0.0
tmp12 = tmp10 <= tmp11
tl.store(out_ptr0 + (x2), tmp10, xmask)
tl.store(out_ptr1 + (x2), tmp12, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9, primals_10, primals_11, primals_12 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_3, (4, 8), (8, 1))
assert_size_stride(primals_4, (4, ), (1, ))
assert_size_stride(primals_5, (4, ), (1, ))
assert_size_stride(primals_6, (4, ), (1, ))
assert_size_stride(primals_7, (4, 4), (4, 1))
assert_size_stride(primals_8, (4, ), (1, ))
assert_size_stride(primals_9, (4, ), (1, ))
assert_size_stride(primals_10, (4, ), (1, ))
assert_size_stride(primals_11, (1, 4), (4, 1))
assert_size_stride(primals_12, (1, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4, 8), (128, 32, 8, 1), torch.float32)
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.cat]
stream0 = get_raw_stream(0)
triton_poi_fused_cat_0.run(primals_1, primals_2, buf0, 512, grid=grid(512), stream=stream0)
del primals_1
del primals_2
buf1 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [linear], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_4, reinterpret_tensor(buf0, (64, 8), (8, 1), 0), reinterpret_tensor(primals_3, (8, 4), (1, 8), 0), alpha=1, beta=1, out=buf1)
del primals_3
del primals_4
buf2 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 64), torch.float32)
buf3 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 64), torch.float32)
# Topologically Sorted Source Nodes: [layer_norm], Original ATen: [aten.native_layer_norm]
triton_poi_fused_native_layer_norm_1.run(buf1, buf2, buf3, 64, grid=grid(64), stream=stream0)
buf4 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
buf12 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.bool)
# Topologically Sorted Source Nodes: [layer_norm, x_1], Original ATen: [aten.native_layer_norm, aten.relu, aten.threshold_backward]
triton_poi_fused_native_layer_norm_relu_threshold_backward_2.run(buf1, buf2, buf3, primals_5, primals_6, buf4, buf12, 256, grid=grid(256), stream=stream0)
del primals_6
buf5 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [linear_1], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_8, reinterpret_tensor(buf4, (64, 4), (4, 1), 0), reinterpret_tensor(primals_7, (4, 4), (1, 4), 0), alpha=1, beta=1, out=buf5)
del primals_8
buf6 = buf3; del buf3 # reuse
buf7 = buf2; del buf2 # reuse
# Topologically Sorted Source Nodes: [layer_norm_1], Original ATen: [aten.native_layer_norm]
triton_poi_fused_native_layer_norm_1.run(buf5, buf6, buf7, 64, grid=grid(64), stream=stream0)
buf8 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
buf11 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.bool)
# Topologically Sorted Source Nodes: [layer_norm_1, x_2], Original ATen: [aten.native_layer_norm, aten.relu, aten.threshold_backward]
triton_poi_fused_native_layer_norm_relu_threshold_backward_2.run(buf5, buf6, buf7, primals_9, primals_10, buf8, buf11, 256, grid=grid(256), stream=stream0)
del buf6
del primals_10
buf10 = reinterpret_tensor(buf7, (64, 1), (1, 1), 0); del buf7 # reuse
# Topologically Sorted Source Nodes: [q], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_12, reinterpret_tensor(buf8, (64, 4), (4, 1), 0), reinterpret_tensor(primals_11, (4, 1), (1, 4), 0), alpha=1, beta=1, out=buf10)
del primals_12
return (reinterpret_tensor(buf10, (4, 4, 4, 1), (16, 4, 1, 1), 0), primals_5, primals_9, reinterpret_tensor(buf0, (64, 8), (8, 1), 0), buf1, reinterpret_tensor(buf4, (64, 4), (4, 1), 0), buf5, reinterpret_tensor(buf8, (64, 4), (4, 1), 0), primals_11, buf11, primals_7, buf12, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 8), (8, 1), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_8 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_9 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_10 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_11 = rand_strided((1, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_12 = rand_strided((1, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9, primals_10, primals_11, primals_12])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch as T
import torch.nn as nn
import torch.optim as optim
class CriticNetwork(nn.Module):
def __init__(self, beta, state_dim, action_dim, fc1_dim, fc2_dim):
super(CriticNetwork, self).__init__()
self.fc1 = nn.Linear(state_dim + action_dim, fc1_dim)
self.ln1 = nn.LayerNorm(fc1_dim)
self.fc2 = nn.Linear(fc1_dim, fc2_dim)
self.ln2 = nn.LayerNorm(fc2_dim)
self.q = nn.Linear(fc2_dim, 1)
self.optimizer = optim.Adam(self.parameters(), lr=beta)
self
def forward(self, state, action):
x = T.cat([state, action], dim=-1)
x = T.relu(self.ln1(self.fc1(x)))
x = T.relu(self.ln2(self.fc2(x)))
q = self.q(x)
return q
def save_checkpoint(self, checkpoint_file):
T.save(self.state_dict(), checkpoint_file,
_use_new_zipfile_serialization=False)
def load_checkpoint(self, checkpoint_file):
self.load_state_dict(T.load(checkpoint_file))
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'beta': 4, 'state_dim': 4, 'action_dim': 4, 'fc1_dim': 4,
'fc2_dim': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import libdevice
import torch as T
import torch.nn as nn
import torch.optim as optim
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_cat_0(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK: tl.
constexpr):
xnumel = 512
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 8
x1 = xindex // 8
x2 = xindex
tmp0 = x0
tl.full([1], 0, tl.int64)
tmp3 = tl.full([1], 4, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = tl.load(in_ptr0 + (4 * x1 + x0), tmp4 & xmask, eviction_policy=
'evict_last', other=0.0)
tmp6 = tmp0 >= tmp3
tl.full([1], 8, tl.int64)
tmp9 = tl.load(in_ptr1 + (4 * x1 + (-4 + x0)), tmp6 & xmask,
eviction_policy='evict_last', other=0.0)
tmp10 = tl.where(tmp4, tmp5, tmp9)
tl.store(out_ptr0 + x2, tmp10, xmask)
@triton.jit
def triton_poi_fused_native_layer_norm_1(in_ptr0, out_ptr0, out_ptr1,
xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + 4 * x0, xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr0 + (1 + 4 * x0), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr0 + (2 + 4 * x0), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (3 + 4 * x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp4 = tmp2 + tmp3
tmp6 = tmp4 + tmp5
tmp7 = 4.0
tmp8 = tmp6 / tmp7
tmp9 = tmp0 - tmp8
tmp10 = tmp9 * tmp9
tmp11 = tmp1 - tmp8
tmp12 = tmp11 * tmp11
tmp13 = tmp10 + tmp12
tmp14 = tmp3 - tmp8
tmp15 = tmp14 * tmp14
tmp16 = tmp13 + tmp15
tmp17 = tmp5 - tmp8
tmp18 = tmp17 * tmp17
tmp19 = tmp16 + tmp18
tmp20 = tmp19 / tmp7
tmp21 = 1e-05
tmp22 = tmp20 + tmp21
tmp23 = libdevice.rsqrt(tmp22)
tl.store(out_ptr0 + x0, tmp8, xmask)
tl.store(out_ptr1 + x0, tmp23, xmask)
@triton.jit
def triton_poi_fused_native_layer_norm_relu_threshold_backward_2(in_ptr0,
in_ptr1, in_ptr2, in_ptr3, in_ptr4, out_ptr0, out_ptr1, xnumel, XBLOCK:
tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
x0 = xindex % 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr1 + x1, xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr2 + x1, xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr3 + x0, xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr4 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 - tmp1
tmp4 = tmp2 * tmp3
tmp6 = tmp4 * tmp5
tmp8 = tmp6 + tmp7
tmp9 = tl.full([1], 0, tl.int32)
tmp10 = triton_helpers.maximum(tmp9, tmp8)
tmp11 = 0.0
tmp12 = tmp10 <= tmp11
tl.store(out_ptr0 + x2, tmp10, xmask)
tl.store(out_ptr1 + x2, tmp12, xmask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7, primals_8, primals_9, primals_10, primals_11, primals_12
) = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_3, (4, 8), (8, 1))
assert_size_stride(primals_4, (4,), (1,))
assert_size_stride(primals_5, (4,), (1,))
assert_size_stride(primals_6, (4,), (1,))
assert_size_stride(primals_7, (4, 4), (4, 1))
assert_size_stride(primals_8, (4,), (1,))
assert_size_stride(primals_9, (4,), (1,))
assert_size_stride(primals_10, (4,), (1,))
assert_size_stride(primals_11, (1, 4), (4, 1))
assert_size_stride(primals_12, (1,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4, 8), (128, 32, 8, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_cat_0[grid(512)](primals_1, primals_2, buf0, 512,
XBLOCK=256, num_warps=4, num_stages=1)
del primals_1
del primals_2
buf1 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
extern_kernels.addmm(primals_4, reinterpret_tensor(buf0, (64, 8), (
8, 1), 0), reinterpret_tensor(primals_3, (8, 4), (1, 8), 0),
alpha=1, beta=1, out=buf1)
del primals_3
del primals_4
buf2 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 64), torch.float32)
buf3 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 64), torch.float32)
triton_poi_fused_native_layer_norm_1[grid(64)](buf1, buf2, buf3, 64,
XBLOCK=64, num_warps=1, num_stages=1)
buf4 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
buf12 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.bool)
triton_poi_fused_native_layer_norm_relu_threshold_backward_2[grid(256)
](buf1, buf2, buf3, primals_5, primals_6, buf4, buf12, 256,
XBLOCK=128, num_warps=4, num_stages=1)
del primals_6
buf5 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
extern_kernels.addmm(primals_8, reinterpret_tensor(buf4, (64, 4), (
4, 1), 0), reinterpret_tensor(primals_7, (4, 4), (1, 4), 0),
alpha=1, beta=1, out=buf5)
del primals_8
buf6 = buf3
del buf3
buf7 = buf2
del buf2
triton_poi_fused_native_layer_norm_1[grid(64)](buf5, buf6, buf7, 64,
XBLOCK=64, num_warps=1, num_stages=1)
buf8 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
buf11 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.bool)
triton_poi_fused_native_layer_norm_relu_threshold_backward_2[grid(256)
](buf5, buf6, buf7, primals_9, primals_10, buf8, buf11, 256,
XBLOCK=128, num_warps=4, num_stages=1)
del buf6
del primals_10
buf10 = reinterpret_tensor(buf7, (64, 1), (1, 1), 0)
del buf7
extern_kernels.addmm(primals_12, reinterpret_tensor(buf8, (64, 4),
(4, 1), 0), reinterpret_tensor(primals_11, (4, 1), (1, 4), 0),
alpha=1, beta=1, out=buf10)
del primals_12
return reinterpret_tensor(buf10, (4, 4, 4, 1), (16, 4, 1, 1), 0
), primals_5, primals_9, reinterpret_tensor(buf0, (64, 8), (8, 1), 0
), buf1, reinterpret_tensor(buf4, (64, 4), (4, 1), 0
), buf5, reinterpret_tensor(buf8, (64, 4), (4, 1), 0
), primals_11, buf11, primals_7, buf12
class CriticNetworkNew(nn.Module):
def __init__(self, beta, state_dim, action_dim, fc1_dim, fc2_dim):
super(CriticNetworkNew, self).__init__()
self.fc1 = nn.Linear(state_dim + action_dim, fc1_dim)
self.ln1 = nn.LayerNorm(fc1_dim)
self.fc2 = nn.Linear(fc1_dim, fc2_dim)
self.ln2 = nn.LayerNorm(fc2_dim)
self.q = nn.Linear(fc2_dim, 1)
self.optimizer = optim.Adam(self.parameters(), lr=beta)
self
def save_checkpoint(self, checkpoint_file):
T.save(self.state_dict(), checkpoint_file,
_use_new_zipfile_serialization=False)
def load_checkpoint(self, checkpoint_file):
self.load_state_dict(T.load(checkpoint_file))
def forward(self, input_0, input_1):
primals_3 = self.fc1.weight
primals_4 = self.fc1.bias
primals_5 = self.ln1.weight
primals_6 = self.ln1.bias
primals_7 = self.fc2.weight
primals_8 = self.fc2.bias
primals_9 = self.ln2.weight
primals_10 = self.ln2.bias
primals_11 = self.q.weight
primals_12 = self.q.bias
primals_1 = input_0
primals_2 = input_1
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7, primals_8, primals_9,
primals_10, primals_11, primals_12])
return output[0]
| indigoLovee/TD3 | CriticNetwork | false | 6,890 | [
"MIT"
] | 1 | 0e86a40c27ec376b52e9f8e0e70db28e7411276b | https://github.com/indigoLovee/TD3/tree/0e86a40c27ec376b52e9f8e0e70db28e7411276b | import torch
import torch as T
import torch.nn as nn
import torch.optim as optim
class Model(nn.Module):
def __init__(self, beta, state_dim, action_dim, fc1_dim, fc2_dim):
super().__init__()
self.fc1 = nn.Linear(state_dim + action_dim, fc1_dim)
self.ln1 = nn.LayerNorm(fc1_dim)
self.fc2 = nn.Linear(fc1_dim, fc2_dim)
self.ln2 = nn.LayerNorm(fc2_dim)
self.q = nn.Linear(fc2_dim, 1)
self.optimizer = optim.Adam(self.parameters(), lr=beta)
self
def forward(self, state, action):
x = T.cat([state, action], dim=-1)
x = T.relu(self.ln1(self.fc1(x)))
x = T.relu(self.ln2(self.fc2(x)))
q = self.q(x)
return q
def save_checkpoint(self, checkpoint_file):
T.save(self.state_dict(), checkpoint_file,
_use_new_zipfile_serialization=False)
def load_checkpoint(self, checkpoint_file):
self.load_state_dict(T.load(checkpoint_file))
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'beta': 4, 'state_dim': 4, 'action_dim': 4, 'fc1_dim': 4,
'fc2_dim': 4}]
|
ChannelSqueeze | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/zb/czbkfibi4iyrs3mosc3eavau6ob7ikc563achj2yzxzrjnqoqfvf.py
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.sum]
# Source node to ATen node mapping:
# x => sum_1
# Graph fragment:
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%view, [2]), kwargs = {})
triton_poi_fused_sum_0 = async_compile.triton('triton_poi_fused_sum_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_sum_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_sum_0(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (x0), xmask)
tl.store(out_ptr0 + (x0), tmp0, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.sum]
stream0 = get_raw_stream(0)
triton_poi_fused_sum_0.run(arg0_1, buf0, 256, grid=grid(256), stream=stream0)
del arg0_1
return (buf0, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([arg0_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
def channel_squeeze(x, groups):
"""
Channel squeeze operation.
Parameters:
----------
x : Tensor
Input tensor.
groups : int
Number of groups.
Returns
-------
Tensor
Resulted tensor.
"""
batch, channels, height, width = x.size()
channels_per_group = channels // groups
x = x.view(batch, channels_per_group, groups, height, width).sum(dim=2)
return x
class ChannelSqueeze(nn.Module):
"""
Channel squeeze layer. This is a wrapper over the same operation. It is designed to save the number of groups.
Parameters:
----------
channels : int
Number of channels.
groups : int
Number of groups.
"""
def __init__(self, channels, groups):
super(ChannelSqueeze, self).__init__()
if channels % groups != 0:
raise ValueError('channels must be divisible by groups')
self.groups = groups
def forward(self, x):
return channel_squeeze(x, self.groups)
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'channels': 4, 'groups': 1}]
| import torch
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_poi_fused_sum_0(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + x0, xmask)
tl.store(out_ptr0 + x0, tmp0, xmask)
def call(args):
arg0_1, = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_sum_0[grid(256)](arg0_1, buf0, 256, XBLOCK=256,
num_warps=4, num_stages=1)
del arg0_1
return buf0,
def channel_squeeze(x, groups):
"""
Channel squeeze operation.
Parameters:
----------
x : Tensor
Input tensor.
groups : int
Number of groups.
Returns
-------
Tensor
Resulted tensor.
"""
batch, channels, height, width = x.size()
channels_per_group = channels // groups
x = x.view(batch, channels_per_group, groups, height, width).sum(dim=2)
return x
class ChannelSqueezeNew(nn.Module):
"""
Channel squeeze layer. This is a wrapper over the same operation. It is designed to save the number of groups.
Parameters:
----------
channels : int
Number of channels.
groups : int
Number of groups.
"""
def __init__(self, channels, groups):
super(ChannelSqueezeNew, self).__init__()
if channels % groups != 0:
raise ValueError('channels must be divisible by groups')
self.groups = groups
def forward(self, input_0):
arg0_1 = input_0
output = call([arg0_1])
return output[0]
| iofthetiger/pkuad | ChannelSqueeze | false | 6,891 | [
"Apache-2.0"
] | 1 | 07496d108c614c84be028f344830becc9cac8fe5 | https://github.com/iofthetiger/pkuad/tree/07496d108c614c84be028f344830becc9cac8fe5 | import torch
import torch.nn as nn
def channel_squeeze(x, groups):
"""
Channel squeeze operation.
Parameters:
----------
x : Tensor
Input tensor.
groups : int
Number of groups.
Returns
-------
Tensor
Resulted tensor.
"""
batch, channels, height, width = x.size()
channels_per_group = channels // groups
x = x.view(batch, channels_per_group, groups, height, width).sum(dim=2)
return x
class Model(nn.Module):
"""
Channel squeeze layer. This is a wrapper over the same operation. It is designed to save the number of groups.
Parameters:
----------
channels : int
Number of channels.
groups : int
Number of groups.
"""
def __init__(self, channels, groups):
super().__init__()
if channels % groups != 0:
raise ValueError('channels must be divisible by groups')
self.groups = groups
def forward(self, x):
return channel_squeeze(x, self.groups)
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4, 1]
|
MP | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/tw/ctwvjh3sk72wvpvfe6qttbbnuv7go7omfvhyfpoli3k62h5jasad.py
# Topologically Sorted Source Nodes: [max_pool2d], Original ATen: [aten.max_pool2d_with_indices]
# Source node to ATen node mapping:
# max_pool2d => getitem
# Graph fragment:
# %getitem : [num_users=1] = call_function[target=operator.getitem](args = (%_low_memory_max_pool2d_with_offsets, 0), kwargs = {})
triton_poi_fused_max_pool2d_with_indices_0 = async_compile.triton('triton_poi_fused_max_pool2d_with_indices_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_max_pool2d_with_indices_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 4, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_max_pool2d_with_indices_0(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 2
x1 = (xindex // 2)
x2 = xindex
tmp0 = tl.load(in_ptr0 + ((2*x0) + (8*x1)), xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr0 + (1 + (2*x0) + (8*x1)), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr0 + (4 + (2*x0) + (8*x1)), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (5 + (2*x0) + (8*x1)), xmask, eviction_policy='evict_last')
tmp2 = triton_helpers.maximum(tmp1, tmp0)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp6 = triton_helpers.maximum(tmp5, tmp4)
tl.store(out_ptr0 + (x2), tmp6, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 2, 2), (16, 4, 2, 1), torch.float32)
# Topologically Sorted Source Nodes: [max_pool2d], Original ATen: [aten.max_pool2d_with_indices]
stream0 = get_raw_stream(0)
triton_poi_fused_max_pool2d_with_indices_0.run(arg0_1, buf0, 64, grid=grid(64), stream=stream0)
del arg0_1
return (buf0, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([arg0_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| from torch.nn import Module
import torch
import torch.utils.data
from torch.nn import MaxPool2d
class MP(Module):
def __init__(self, k=2):
super().__init__()
self.m = MaxPool2d(kernel_size=k, stride=k)
def forward(self, x):
return self.m(x)
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {}]
| import torch
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch.nn import Module
import torch.utils.data
from torch.nn import MaxPool2d
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_poi_fused_max_pool2d_with_indices_0(in_ptr0, out_ptr0, xnumel,
XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 2
x1 = xindex // 2
x2 = xindex
tmp0 = tl.load(in_ptr0 + (2 * x0 + 8 * x1), xmask, eviction_policy=
'evict_last')
tmp1 = tl.load(in_ptr0 + (1 + 2 * x0 + 8 * x1), xmask, eviction_policy=
'evict_last')
tmp3 = tl.load(in_ptr0 + (4 + 2 * x0 + 8 * x1), xmask, eviction_policy=
'evict_last')
tmp5 = tl.load(in_ptr0 + (5 + 2 * x0 + 8 * x1), xmask, eviction_policy=
'evict_last')
tmp2 = triton_helpers.maximum(tmp1, tmp0)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp6 = triton_helpers.maximum(tmp5, tmp4)
tl.store(out_ptr0 + x2, tmp6, xmask)
def call(args):
arg0_1, = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 2, 2), (16, 4, 2, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_max_pool2d_with_indices_0[grid(64)](arg0_1, buf0,
64, XBLOCK=64, num_warps=1, num_stages=1)
del arg0_1
return buf0,
class MPNew(Module):
def __init__(self, k=2):
super().__init__()
self.m = MaxPool2d(kernel_size=k, stride=k)
def forward(self, input_0):
arg0_1 = input_0
output = call([arg0_1])
return output[0]
| ioangatop/yolo | MP | false | 6,892 | [
"MIT"
] | 1 | c65a72337369572bc07090f39123e2bf6ff5f4a3 | https://github.com/ioangatop/yolo/tree/c65a72337369572bc07090f39123e2bf6ff5f4a3 | from torch.nn import Module
import torch
import torch.utils.data
from torch.nn import MaxPool2d
class Model(Module):
def __init__(self, k=2):
super().__init__()
self.m = MaxPool2d(kernel_size=k, stride=k)
def forward(self, x):
return self.m(x)
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return []
|
BasicBlockWN | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/p7/cp7mjs4cvfmmk3xanldfhbfyw3iz6jtioejhqhvyiw5cvteee5uz.py
# Topologically Sorted Source Nodes: [_weight_norm], Original ATen: [aten._weight_norm_interface]
# Source node to ATen node mapping:
# _weight_norm => div, mul, pow_1, pow_2, sum_1
# Graph fragment:
# %pow_1 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%primals_2, 2), kwargs = {})
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%pow_1, [1, 2, 3], True), kwargs = {})
# %pow_2 : [num_users=2] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%sum_1, 0.5), kwargs = {})
# %div : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%primals_1, %pow_2), kwargs = {})
# %mul : [num_users=2] = call_function[target=torch.ops.aten.mul.Tensor](args = (%primals_2, %div), kwargs = {})
triton_per_fused__weight_norm_interface_0 = async_compile.triton('triton_per_fused__weight_norm_interface_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[4, 64],
reduction_hint=ReductionHint.INNER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32', 5: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused__weight_norm_interface_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 1, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused__weight_norm_interface_0(in_out_ptr0, in_ptr0, in_ptr1, out_ptr0, xnumel, rnumel, XBLOCK : tl.constexpr):
xnumel = 4
rnumel = 36
RBLOCK: tl.constexpr = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
roffset = 0
rmask = rindex < rnumel
r1 = rindex
x0 = xindex
tmp0 = tl.load(in_ptr0 + (r1 + (36*x0)), rmask & xmask, other=0.0)
tmp7 = tl.load(in_ptr1 + (x0), xmask, eviction_policy='evict_last')
tmp1 = tmp0 * tmp0
tmp2 = tl.broadcast_to(tmp1, [XBLOCK, RBLOCK])
tmp4 = tl.where(rmask & xmask, tmp2, 0)
tmp5 = tl.sum(tmp4, 1)[:, None]
tmp6 = libdevice.sqrt(tmp5)
tmp8 = tmp7 / tmp6
tmp9 = tmp0 * tmp8
tl.debug_barrier()
tl.store(in_out_ptr0 + (x0), tmp6, xmask)
tl.store(out_ptr0 + (r1 + (36*x0)), tmp9, rmask & xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/3g/c3gulbvr4xrfq3wps6kqjc3yuakrgtdcdvb44tmfrvggj56xwcm6.py
# Topologically Sorted Source Nodes: [out], Original ATen: [aten.relu]
# Source node to ATen node mapping:
# out => relu
# Graph fragment:
# %relu : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%convolution,), kwargs = {})
triton_poi_fused_relu_1 = async_compile.triton('triton_poi_fused_relu_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_relu_1', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_relu_1(in_out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_out_ptr0 + (x0), xmask)
tmp1 = tl.full([1], 0, tl.int32)
tmp2 = triton_helpers.maximum(tmp1, tmp0)
tl.store(in_out_ptr0 + (x0), tmp2, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/et/cet7mpe46xrhpaxasnjmwa5kqarykvc3xmaemz25bq4hpkn3n7cb.py
# Topologically Sorted Source Nodes: [out_2, out_3], Original ATen: [aten.add, aten.relu, aten.threshold_backward]
# Source node to ATen node mapping:
# out_2 => add
# out_3 => relu_1
# Graph fragment:
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%convolution_1, %primals_3), kwargs = {})
# %relu_1 : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%add,), kwargs = {})
# %le : [num_users=1] = call_function[target=torch.ops.aten.le.Scalar](args = (%relu_1, 0), kwargs = {})
triton_poi_fused_add_relu_threshold_backward_2 = async_compile.triton('triton_poi_fused_add_relu_threshold_backward_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*i1', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_relu_threshold_backward_2', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_relu_threshold_backward_2(in_out_ptr0, in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_out_ptr0 + (x0), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask)
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(in_out_ptr0 + (x0), tmp4, xmask)
tl.store(out_ptr0 + (x0), tmp6, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5 = args
args.clear()
assert_size_stride(primals_1, (4, 1, 1, 1), (1, 1, 1, 1))
assert_size_stride(primals_2, (4, 4, 3, 3), (36, 9, 3, 1))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_4, (4, 1, 1, 1), (1, 1, 1, 1))
assert_size_stride(primals_5, (4, 4, 3, 3), (36, 9, 3, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 1, 1, 1), (1, 4, 4, 4), torch.float32)
buf1 = reinterpret_tensor(buf0, (4, 1, 1, 1), (1, 1, 1, 1), 0); del buf0 # reuse
buf2 = empty_strided_cuda((4, 4, 3, 3), (36, 9, 3, 1), torch.float32)
# Topologically Sorted Source Nodes: [_weight_norm], Original ATen: [aten._weight_norm_interface]
stream0 = get_raw_stream(0)
triton_per_fused__weight_norm_interface_0.run(buf1, primals_2, primals_1, buf2, 4, 36, grid=grid(4), stream=stream0)
# Topologically Sorted Source Nodes: [conv2d], Original ATen: [aten.convolution]
buf3 = extern_kernels.convolution(primals_3, buf2, stride=(1, 1), padding=(1, 1), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf3, (4, 4, 4, 4), (64, 16, 4, 1))
buf4 = buf3; del buf3 # reuse
# Topologically Sorted Source Nodes: [out], Original ATen: [aten.relu]
triton_poi_fused_relu_1.run(buf4, 256, grid=grid(256), stream=stream0)
buf5 = empty_strided_cuda((4, 1, 1, 1), (1, 4, 4, 4), torch.float32)
buf6 = reinterpret_tensor(buf5, (4, 1, 1, 1), (1, 1, 1, 1), 0); del buf5 # reuse
buf7 = empty_strided_cuda((4, 4, 3, 3), (36, 9, 3, 1), torch.float32)
# Topologically Sorted Source Nodes: [_weight_norm_1], Original ATen: [aten._weight_norm_interface]
triton_per_fused__weight_norm_interface_0.run(buf6, primals_5, primals_4, buf7, 4, 36, grid=grid(4), stream=stream0)
# Topologically Sorted Source Nodes: [out_1], Original ATen: [aten.convolution]
buf8 = extern_kernels.convolution(buf4, buf7, stride=(1, 1), padding=(1, 1), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf8, (4, 4, 4, 4), (64, 16, 4, 1))
buf9 = buf8; del buf8 # reuse
buf10 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.bool)
# Topologically Sorted Source Nodes: [out_2, out_3], Original ATen: [aten.add, aten.relu, aten.threshold_backward]
triton_poi_fused_add_relu_threshold_backward_2.run(buf9, primals_3, buf10, 256, grid=grid(256), stream=stream0)
return (buf9, buf2, buf7, primals_1, primals_2, primals_3, primals_4, primals_5, buf1, buf2, buf4, buf6, buf7, buf10, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 1, 1, 1), (1, 1, 1, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 4, 3, 3), (36, 9, 3, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, 1, 1, 1), (1, 1, 1, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((4, 4, 3, 3), (36, 9, 3, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch as t
import torch.nn as nn
from abc import ABC
from torch.nn.utils.weight_norm import weight_norm
def conv1x1(in_planes, out_planes, stride=1):
"""
Create a 1x1 2d convolution block
"""
return nn.Conv2d(in_planes, out_planes, kernel_size=1, stride=stride,
bias=False)
def conv3x3(in_planes, out_planes, stride=1):
"""
Create a 3x3 2d convolution block
"""
return nn.Conv2d(in_planes, out_planes, kernel_size=3, stride=stride,
padding=1, bias=False)
class NeuralNetworkModule(nn.Module, ABC):
"""
Note: input device and output device are determined by module parameters,
your input module / output submodule should not store parameters on
more than one device, and you also should not move your output to
other devices other than your parameter storage device in forward().
"""
def __init__(self):
super().__init__()
self.input_module = None
self.output_module = None
def set_input_module(self, input_module: 'nn.Module'):
"""
Set the input submodule of current module.
"""
self.input_module = input_module
if not isinstance(input_module, NeuralNetworkModule):
if isinstance(input_module, nn.Sequential):
input_module = self.find_child(input_module, True)
if len({p.device for p in input_module.parameters()}) > 1:
raise RuntimeError(
'Input module must be another NeuralNetworkModule or locate on one single device.'
)
def set_output_module(self, output_module: 'nn.Module'):
"""
Set the output submodule of current module.
"""
self.output_module = output_module
if not isinstance(output_module, NeuralNetworkModule):
if isinstance(output_module, nn.Sequential):
output_module = self.find_child(output_module, False)
if len({p.device for p in output_module.parameters()}) > 1:
raise RuntimeError(
'Output module must be another NeuralNetworkModule or locate on one single device.'
)
@property
def input_device(self):
if self.input_module is None:
raise RuntimeError('Input module not set.')
elif not isinstance(self.input_module, NeuralNetworkModule):
dev_set = {p.device for p in self.input_module.parameters()}
if len(dev_set) != 1:
raise RuntimeError(
'This input module contains parameters on different devices, please consider about splitting it.'
)
else:
return list(dev_set)[0]
else:
return self.input_module.input_device
@property
def output_device(self):
if self.output_module is None and self.input_module is None:
raise RuntimeError('Output module not set.')
elif self.output_module is not None:
if not isinstance(self.output_module, NeuralNetworkModule):
dev_set = {p.device for p in self.output_module.parameters()}
if len(dev_set) != 1:
raise RuntimeError(
'This output module contains parameters on different devices, please consider about splitting it.'
)
else:
return list(dev_set)[0]
else:
return self.output_module.output_device
else:
return self.input_device
@staticmethod
def find_child(seq, is_first=True):
"""
Find the first / last leaf child module.
"""
if isinstance(seq, nn.Sequential):
if is_first:
return NeuralNetworkModule.find_child(seq[0], is_first)
else:
return NeuralNetworkModule.find_child(seq[-1], is_first)
else:
return seq
def forward(self, *_):
pass
class BasicBlockWN(NeuralNetworkModule):
"""
Basic block with weight normalization
"""
expansion = 1
def __init__(self, in_planes, out_planes, stride=1, **__):
"""
Create a basic block of resnet.
Args:
in_planes: Number of input planes.
out_planes: Number of output planes.
stride: Stride of convolution.
"""
super().__init__()
self.conv1 = weight_norm(conv3x3(in_planes, out_planes, stride))
self.conv2 = weight_norm(conv3x3(out_planes, self.expansion *
out_planes))
self.shortcut = nn.Sequential()
self.set_input_module(self.conv1)
if stride != 1 or in_planes != self.expansion * out_planes:
self.shortcut = nn.Sequential(weight_norm(conv1x1(in_planes,
self.expansion * out_planes, stride)))
def forward(self, x):
out = t.relu(self.conv1(x))
out = self.conv2(out)
out += self.shortcut(x)
out = t.relu(out)
return out
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'in_planes': 4, 'out_planes': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import libdevice
import torch.nn as nn
from abc import ABC
from torch.nn.utils.weight_norm import weight_norm
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_per_fused__weight_norm_interface_0(in_out_ptr0, in_ptr0, in_ptr1,
out_ptr0, xnumel, rnumel, XBLOCK: tl.constexpr):
xnumel = 4
rnumel = 36
RBLOCK: tl.constexpr = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
rmask = rindex < rnumel
r1 = rindex
x0 = xindex
tmp0 = tl.load(in_ptr0 + (r1 + 36 * x0), rmask & xmask, other=0.0)
tmp7 = tl.load(in_ptr1 + x0, xmask, eviction_policy='evict_last')
tmp1 = tmp0 * tmp0
tmp2 = tl.broadcast_to(tmp1, [XBLOCK, RBLOCK])
tmp4 = tl.where(rmask & xmask, tmp2, 0)
tmp5 = tl.sum(tmp4, 1)[:, None]
tmp6 = libdevice.sqrt(tmp5)
tmp8 = tmp7 / tmp6
tmp9 = tmp0 * tmp8
tl.debug_barrier()
tl.store(in_out_ptr0 + x0, tmp6, xmask)
tl.store(out_ptr0 + (r1 + 36 * x0), tmp9, rmask & xmask)
@triton.jit
def triton_poi_fused_relu_1(in_out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_out_ptr0 + x0, xmask)
tmp1 = tl.full([1], 0, tl.int32)
tmp2 = triton_helpers.maximum(tmp1, tmp0)
tl.store(in_out_ptr0 + x0, tmp2, xmask)
@triton.jit
def triton_poi_fused_add_relu_threshold_backward_2(in_out_ptr0, in_ptr0,
out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_out_ptr0 + x0, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask)
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(in_out_ptr0 + x0, tmp4, xmask)
tl.store(out_ptr0 + x0, tmp6, xmask)
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5 = args
args.clear()
assert_size_stride(primals_1, (4, 1, 1, 1), (1, 1, 1, 1))
assert_size_stride(primals_2, (4, 4, 3, 3), (36, 9, 3, 1))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_4, (4, 1, 1, 1), (1, 1, 1, 1))
assert_size_stride(primals_5, (4, 4, 3, 3), (36, 9, 3, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 1, 1, 1), (1, 4, 4, 4), torch.float32)
buf1 = reinterpret_tensor(buf0, (4, 1, 1, 1), (1, 1, 1, 1), 0)
del buf0
buf2 = empty_strided_cuda((4, 4, 3, 3), (36, 9, 3, 1), torch.float32)
get_raw_stream(0)
triton_per_fused__weight_norm_interface_0[grid(4)](buf1, primals_2,
primals_1, buf2, 4, 36, XBLOCK=1, num_warps=2, num_stages=1)
buf3 = extern_kernels.convolution(primals_3, buf2, stride=(1, 1),
padding=(1, 1), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf3, (4, 4, 4, 4), (64, 16, 4, 1))
buf4 = buf3
del buf3
triton_poi_fused_relu_1[grid(256)](buf4, 256, XBLOCK=128, num_warps
=4, num_stages=1)
buf5 = empty_strided_cuda((4, 1, 1, 1), (1, 4, 4, 4), torch.float32)
buf6 = reinterpret_tensor(buf5, (4, 1, 1, 1), (1, 1, 1, 1), 0)
del buf5
buf7 = empty_strided_cuda((4, 4, 3, 3), (36, 9, 3, 1), torch.float32)
triton_per_fused__weight_norm_interface_0[grid(4)](buf6, primals_5,
primals_4, buf7, 4, 36, XBLOCK=1, num_warps=2, num_stages=1)
buf8 = extern_kernels.convolution(buf4, buf7, stride=(1, 1),
padding=(1, 1), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf8, (4, 4, 4, 4), (64, 16, 4, 1))
buf9 = buf8
del buf8
buf10 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.bool)
triton_poi_fused_add_relu_threshold_backward_2[grid(256)](buf9,
primals_3, buf10, 256, XBLOCK=128, num_warps=4, num_stages=1)
return (buf9, buf2, buf7, primals_1, primals_2, primals_3, primals_4,
primals_5, buf1, buf2, buf4, buf6, buf7, buf10)
def conv1x1(in_planes, out_planes, stride=1):
"""
Create a 1x1 2d convolution block
"""
return nn.Conv2d(in_planes, out_planes, kernel_size=1, stride=stride,
bias=False)
def conv3x3(in_planes, out_planes, stride=1):
"""
Create a 3x3 2d convolution block
"""
return nn.Conv2d(in_planes, out_planes, kernel_size=3, stride=stride,
padding=1, bias=False)
class NeuralNetworkModule(nn.Module, ABC):
"""
Note: input device and output device are determined by module parameters,
your input module / output submodule should not store parameters on
more than one device, and you also should not move your output to
other devices other than your parameter storage device in forward().
"""
def __init__(self):
super().__init__()
self.input_module = None
self.output_module = None
def set_input_module(self, input_module: 'nn.Module'):
"""
Set the input submodule of current module.
"""
self.input_module = input_module
if not isinstance(input_module, NeuralNetworkModule):
if isinstance(input_module, nn.Sequential):
input_module = self.find_child(input_module, True)
if len({p.device for p in input_module.parameters()}) > 1:
raise RuntimeError(
'Input module must be another NeuralNetworkModule or locate on one single device.'
)
def set_output_module(self, output_module: 'nn.Module'):
"""
Set the output submodule of current module.
"""
self.output_module = output_module
if not isinstance(output_module, NeuralNetworkModule):
if isinstance(output_module, nn.Sequential):
output_module = self.find_child(output_module, False)
if len({p.device for p in output_module.parameters()}) > 1:
raise RuntimeError(
'Output module must be another NeuralNetworkModule or locate on one single device.'
)
@property
def input_device(self):
if self.input_module is None:
raise RuntimeError('Input module not set.')
elif not isinstance(self.input_module, NeuralNetworkModule):
dev_set = {p.device for p in self.input_module.parameters()}
if len(dev_set) != 1:
raise RuntimeError(
'This input module contains parameters on different devices, please consider about splitting it.'
)
else:
return list(dev_set)[0]
else:
return self.input_module.input_device
@property
def output_device(self):
if self.output_module is None and self.input_module is None:
raise RuntimeError('Output module not set.')
elif self.output_module is not None:
if not isinstance(self.output_module, NeuralNetworkModule):
dev_set = {p.device for p in self.output_module.parameters()}
if len(dev_set) != 1:
raise RuntimeError(
'This output module contains parameters on different devices, please consider about splitting it.'
)
else:
return list(dev_set)[0]
else:
return self.output_module.output_device
else:
return self.input_device
@staticmethod
def find_child(seq, is_first=True):
"""
Find the first / last leaf child module.
"""
if isinstance(seq, nn.Sequential):
if is_first:
return NeuralNetworkModule.find_child(seq[0], is_first)
else:
return NeuralNetworkModule.find_child(seq[-1], is_first)
else:
return seq
def forward(self, *_):
pass
class BasicBlockWNNew(NeuralNetworkModule):
"""
Basic block with weight normalization
"""
expansion = 1
def __init__(self, in_planes, out_planes, stride=1, **__):
"""
Create a basic block of resnet.
Args:
in_planes: Number of input planes.
out_planes: Number of output planes.
stride: Stride of convolution.
"""
super().__init__()
self.conv1 = weight_norm(conv3x3(in_planes, out_planes, stride))
self.conv2 = weight_norm(conv3x3(out_planes, self.expansion *
out_planes))
self.shortcut = nn.Sequential()
self.set_input_module(self.conv1)
if stride != 1 or in_planes != self.expansion * out_planes:
self.shortcut = nn.Sequential(weight_norm(conv1x1(in_planes,
self.expansion * out_planes, stride)))
def forward(self, input_0):
primals_1 = self.conv1.weight_g
primals_2 = self.conv1.weight_v
primals_4 = self.conv2.weight_g
primals_5 = self.conv2.weight_v
primals_3 = input_0
output = call([primals_1, primals_2, primals_3, primals_4, primals_5])
return output[0]
| ikamensh/machin | BasicBlockWN | false | 6,893 | [
"MIT"
] | 1 | af7b423c47bc1412530cf6c96c11bd3af9b3e239 | https://github.com/ikamensh/machin/tree/af7b423c47bc1412530cf6c96c11bd3af9b3e239 | import torch
import torch as t
import torch.nn as nn
from abc import ABC
from torch.nn.utils.weight_norm import weight_norm
def conv1x1(in_planes, out_planes, stride=1):
"""
Create a 1x1 2d convolution block
"""
return nn.Conv2d(in_planes, out_planes, kernel_size=1, stride=stride,
bias=False)
def conv3x3(in_planes, out_planes, stride=1):
"""
Create a 3x3 2d convolution block
"""
return nn.Conv2d(in_planes, out_planes, kernel_size=3, stride=stride,
padding=1, bias=False)
class NeuralNetworkModule(nn.Module, ABC):
"""
Note: input device and output device are determined by module parameters,
your input module / output submodule should not store parameters on
more than one device, and you also should not move your output to
other devices other than your parameter storage device in forward().
"""
def __init__(self):
super().__init__()
self.input_module = None
self.output_module = None
def set_input_module(self, input_module: 'nn.Module'):
"""
Set the input submodule of current module.
"""
self.input_module = input_module
if not isinstance(input_module, NeuralNetworkModule):
if isinstance(input_module, nn.Sequential):
input_module = self.find_child(input_module, True)
if len({p.device for p in input_module.parameters()}) > 1:
raise RuntimeError(
'Input module must be another NeuralNetworkModule or locate on one single device.'
)
def set_output_module(self, output_module: 'nn.Module'):
"""
Set the output submodule of current module.
"""
self.output_module = output_module
if not isinstance(output_module, NeuralNetworkModule):
if isinstance(output_module, nn.Sequential):
output_module = self.find_child(output_module, False)
if len({p.device for p in output_module.parameters()}) > 1:
raise RuntimeError(
'Output module must be another NeuralNetworkModule or locate on one single device.'
)
@property
def input_device(self):
if self.input_module is None:
raise RuntimeError('Input module not set.')
elif not isinstance(self.input_module, NeuralNetworkModule):
dev_set = {p.device for p in self.input_module.parameters()}
if len(dev_set) != 1:
raise RuntimeError(
'This input module contains parameters on different devices, please consider about splitting it.'
)
else:
return list(dev_set)[0]
else:
return self.input_module.input_device
@property
def output_device(self):
if self.output_module is None and self.input_module is None:
raise RuntimeError('Output module not set.')
elif self.output_module is not None:
if not isinstance(self.output_module, NeuralNetworkModule):
dev_set = {p.device for p in self.output_module.parameters()}
if len(dev_set) != 1:
raise RuntimeError(
'This output module contains parameters on different devices, please consider about splitting it.'
)
else:
return list(dev_set)[0]
else:
return self.output_module.output_device
else:
return self.input_device
@staticmethod
def find_child(seq, is_first=True):
"""
Find the first / last leaf child module.
"""
if isinstance(seq, nn.Sequential):
if is_first:
return NeuralNetworkModule.find_child(seq[0], is_first)
else:
return NeuralNetworkModule.find_child(seq[-1], is_first)
else:
r
# ... truncated (>4000 chars) for memory efficiency |
Gated_Conv_1d | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/jd/cjdqbkykpbat77atzojktncufemrhk3jioq3xwatv2sxgl7uebzv.py
# Topologically Sorted Source Nodes: [tanh, sigmoid, mul], Original ATen: [aten.tanh, aten.sigmoid, aten.mul]
# Source node to ATen node mapping:
# mul => mul
# sigmoid => sigmoid
# tanh => tanh
# Graph fragment:
# %tanh : [num_users=2] = call_function[target=torch.ops.aten.tanh.default](args = (%getitem,), kwargs = {})
# %sigmoid : [num_users=2] = call_function[target=torch.ops.aten.sigmoid.default](args = (%getitem_1,), kwargs = {})
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%tanh, %sigmoid), kwargs = {})
triton_poi_fused_mul_sigmoid_tanh_0 = async_compile.triton('triton_poi_fused_mul_sigmoid_tanh_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: '*fp32', 5: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_mul_sigmoid_tanh_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 4, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_mul_sigmoid_tanh_0(in_ptr0, in_ptr1, out_ptr0, out_ptr1, out_ptr2, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 4
x1 = (xindex // 4)
x2 = xindex
tmp0 = tl.load(in_ptr0 + (x0 + (8*x1)), xmask)
tmp1 = tl.load(in_ptr1 + (x0), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (4 + x0 + (8*x1)), xmask)
tmp5 = tl.load(in_ptr1 + (4 + x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = libdevice.tanh(tmp2)
tmp6 = tmp4 + tmp5
tmp7 = tl.sigmoid(tmp6)
tmp8 = tmp3 * tmp7
tl.store(out_ptr0 + (x2), tmp3, xmask)
tl.store(out_ptr1 + (x2), tmp7, xmask)
tl.store(out_ptr2 + (x2), tmp8, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (8, 4, 4), (16, 4, 1))
assert_size_stride(primals_2, (8, ), (1, ))
assert_size_stride(primals_3, (4, 4, 4), (16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.convolution]
buf0 = extern_kernels.convolution(primals_3, primals_1, stride=(1,), padding=(0,), dilation=(1,), transposed=False, output_padding=(0,), groups=1, bias=None)
assert_size_stride(buf0, (4, 8, 1), (8, 1, 1))
buf1 = empty_strided_cuda((4, 4, 1), (4, 1, 1), torch.float32)
buf2 = empty_strided_cuda((4, 4, 1), (4, 1, 1), torch.float32)
buf3 = empty_strided_cuda((4, 4, 1), (4, 1, 1), torch.float32)
# Topologically Sorted Source Nodes: [tanh, sigmoid, mul], Original ATen: [aten.tanh, aten.sigmoid, aten.mul]
stream0 = get_raw_stream(0)
triton_poi_fused_mul_sigmoid_tanh_0.run(buf0, primals_2, buf1, buf2, buf3, 16, grid=grid(16), stream=stream0)
del buf0
del primals_2
return (buf3, primals_1, primals_3, buf1, buf2, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((8, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((8, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
class Gated_Conv_1d(nn.Module):
def __init__(self, channels, kernel_size, stride=1, padding=0, dilation
=1, groups=1, bias=True):
super(Gated_Conv_1d, self).__init__()
self.dilation = dilation
self.channels = channels
self.conv_dil = nn.Conv1d(in_channels=channels, out_channels=2 *
channels, kernel_size=kernel_size, stride=stride, padding=
padding, dilation=dilation, groups=groups, bias=bias)
self.tan = nn.Tanh()
self.sig = nn.Sigmoid()
def forward(self, x):
x = self.conv_dil(x)
tn, sg = torch.split(x, self.channels, 1)
return self.tan(tn) * self.sig(sg)
def get_inputs():
return [torch.rand([4, 4, 4])]
def get_init_inputs():
return [[], {'channels': 4, 'kernel_size': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime.triton_helpers import libdevice
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_poi_fused_mul_sigmoid_tanh_0(in_ptr0, in_ptr1, out_ptr0,
out_ptr1, out_ptr2, xnumel, XBLOCK: tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 4
x1 = xindex // 4
x2 = xindex
tmp0 = tl.load(in_ptr0 + (x0 + 8 * x1), xmask)
tmp1 = tl.load(in_ptr1 + x0, xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (4 + x0 + 8 * x1), xmask)
tmp5 = tl.load(in_ptr1 + (4 + x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = libdevice.tanh(tmp2)
tmp6 = tmp4 + tmp5
tmp7 = tl.sigmoid(tmp6)
tmp8 = tmp3 * tmp7
tl.store(out_ptr0 + x2, tmp3, xmask)
tl.store(out_ptr1 + x2, tmp7, xmask)
tl.store(out_ptr2 + x2, tmp8, xmask)
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (8, 4, 4), (16, 4, 1))
assert_size_stride(primals_2, (8,), (1,))
assert_size_stride(primals_3, (4, 4, 4), (16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = extern_kernels.convolution(primals_3, primals_1, stride=(1,),
padding=(0,), dilation=(1,), transposed=False, output_padding=(
0,), groups=1, bias=None)
assert_size_stride(buf0, (4, 8, 1), (8, 1, 1))
buf1 = empty_strided_cuda((4, 4, 1), (4, 1, 1), torch.float32)
buf2 = empty_strided_cuda((4, 4, 1), (4, 1, 1), torch.float32)
buf3 = empty_strided_cuda((4, 4, 1), (4, 1, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_mul_sigmoid_tanh_0[grid(16)](buf0, primals_2, buf1,
buf2, buf3, 16, XBLOCK=16, num_warps=1, num_stages=1)
del buf0
del primals_2
return buf3, primals_1, primals_3, buf1, buf2
class Gated_Conv_1dNew(nn.Module):
def __init__(self, channels, kernel_size, stride=1, padding=0, dilation
=1, groups=1, bias=True):
super(Gated_Conv_1dNew, self).__init__()
self.dilation = dilation
self.channels = channels
self.conv_dil = nn.Conv1d(in_channels=channels, out_channels=2 *
channels, kernel_size=kernel_size, stride=stride, padding=
padding, dilation=dilation, groups=groups, bias=bias)
self.tan = nn.Tanh()
self.sig = nn.Sigmoid()
def forward(self, input_0):
primals_1 = self.conv_dil.weight
primals_2 = self.conv_dil.bias
primals_3 = input_0
output = call([primals_1, primals_2, primals_3])
return output[0]
| ioanvl/wavenet_classifier_torch | Gated_Conv_1d | false | 6,894 | [
"MIT"
] | 1 | de29bfce59d52ae46143f62c4d7a6158a04edf00 | https://github.com/ioanvl/wavenet_classifier_torch/tree/de29bfce59d52ae46143f62c4d7a6158a04edf00 | import torch
import torch.nn as nn
class Model(nn.Module):
def __init__(self, channels, kernel_size, stride=1, padding=0, dilation
=1, groups=1, bias=True):
super().__init__()
self.dilation = dilation
self.channels = channels
self.conv_dil = nn.Conv1d(in_channels=channels, out_channels=2 *
channels, kernel_size=kernel_size, stride=stride, padding=
padding, dilation=dilation, groups=groups, bias=bias)
self.tan = nn.Tanh()
self.sig = nn.Sigmoid()
def forward(self, x):
x = self.conv_dil(x)
tn, sg = torch.split(x, self.channels, 1)
return self.tan(tn) * self.sig(sg)
def get_inputs():
return [torch.rand([4, 4, 4])]
def get_init_inputs():
return [4, 4]
|
IRevInjectivePad | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/ct/cct5cn6mo6zny2z4rzeplk2fltbb4cibi5gdqoa7cpdxa6kif7lx.py
# Topologically Sorted Source Nodes: [x_1], Original ATen: [aten.constant_pad_nd]
# Source node to ATen node mapping:
# x_1 => constant_pad_nd
# Graph fragment:
# %constant_pad_nd : [num_users=1] = call_function[target=torch.ops.aten.constant_pad_nd.default](args = (%permute, [0, 0, 0, 4], 0.0), kwargs = {})
triton_poi_fused_constant_pad_nd_0 = async_compile.triton('triton_poi_fused_constant_pad_nd_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[512],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_constant_pad_nd_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_constant_pad_nd_0(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 512
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = (xindex // 4) % 8
x0 = xindex % 4
x2 = (xindex // 32) % 4
x3 = (xindex // 128)
x4 = xindex
tmp0 = x1
tmp1 = tl.full([1], 4, tl.int64)
tmp2 = tmp0 < tmp1
tmp3 = tl.load(in_ptr0 + (x0 + (4*x2) + (16*x1) + (64*x3)), tmp2 & xmask, other=0.0)
tl.store(out_ptr0 + (x4), tmp3, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 8, 4), (128, 32, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [x_1], Original ATen: [aten.constant_pad_nd]
stream0 = get_raw_stream(0)
triton_poi_fused_constant_pad_nd_0.run(arg0_1, buf0, 512, grid=grid(512), stream=stream0)
del arg0_1
return (reinterpret_tensor(buf0, (4, 8, 4, 4), (128, 4, 32, 1), 0), )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([arg0_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
class IRevInjectivePad(nn.Module):
"""
i-RevNet channel zero padding block.
Parameters:
----------
padding : int
Size of the padding.
"""
def __init__(self, padding):
super(IRevInjectivePad, self).__init__()
self.padding = padding
self.pad = nn.ZeroPad2d(padding=(0, 0, 0, padding))
def forward(self, x):
x = x.permute(0, 2, 1, 3)
x = self.pad(x)
return x.permute(0, 2, 1, 3)
def inverse(self, x):
return x[:, :x.size(1) - self.padding, :, :]
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'padding': 4}]
| import torch
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_constant_pad_nd_0(in_ptr0, out_ptr0, xnumel, XBLOCK:
tl.constexpr):
xnumel = 512
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = xindex // 4 % 8
x0 = xindex % 4
x2 = xindex // 32 % 4
x3 = xindex // 128
x4 = xindex
tmp0 = x1
tmp1 = tl.full([1], 4, tl.int64)
tmp2 = tmp0 < tmp1
tmp3 = tl.load(in_ptr0 + (x0 + 4 * x2 + 16 * x1 + 64 * x3), tmp2 &
xmask, other=0.0)
tl.store(out_ptr0 + x4, tmp3, xmask)
def call(args):
arg0_1, = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 8, 4), (128, 32, 4, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_constant_pad_nd_0[grid(512)](arg0_1, buf0, 512,
XBLOCK=256, num_warps=4, num_stages=1)
del arg0_1
return reinterpret_tensor(buf0, (4, 8, 4, 4), (128, 4, 32, 1), 0),
class IRevInjectivePadNew(nn.Module):
"""
i-RevNet channel zero padding block.
Parameters:
----------
padding : int
Size of the padding.
"""
def __init__(self, padding):
super(IRevInjectivePadNew, self).__init__()
self.padding = padding
self.pad = nn.ZeroPad2d(padding=(0, 0, 0, padding))
def inverse(self, x):
return x[:, :x.size(1) - self.padding, :, :]
def forward(self, input_0):
arg0_1 = input_0
output = call([arg0_1])
return output[0]
| iofthetiger/pkuad | IRevInjectivePad | false | 6,895 | [
"Apache-2.0"
] | 1 | 07496d108c614c84be028f344830becc9cac8fe5 | https://github.com/iofthetiger/pkuad/tree/07496d108c614c84be028f344830becc9cac8fe5 | import torch
import torch.nn as nn
class Model(nn.Module):
"""
i-RevNet channel zero padding block.
Parameters:
----------
padding : int
Size of the padding.
"""
def __init__(self, padding):
super().__init__()
self.padding = padding
self.pad = nn.ZeroPad2d(padding=(0, 0, 0, padding))
def forward(self, x):
x = x.permute(0, 2, 1, 3)
x = self.pad(x)
return x.permute(0, 2, 1, 3)
def inverse(self, x):
return x[:, :x.size(1) - self.padding, :, :]
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4]
|
BERTIntermediate | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/6s/c6shmuvjmq6zc4ifvdsynorwri47ra63qxa7jg3e7p6lw6xlqj5q.py
# Topologically Sorted Source Nodes: [mul, truediv, erf, add, hidden_states_1], Original ATen: [aten.mul, aten.div, aten.erf, aten.add]
# Source node to ATen node mapping:
# add => add
# erf => erf
# hidden_states_1 => mul_1
# mul => mul
# truediv => div
# Graph fragment:
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%view_1, 0.5), kwargs = {})
# %div : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%view_1, 1.4142135623730951), kwargs = {})
# %erf : [num_users=1] = call_function[target=torch.ops.aten.erf.default](args = (%div,), kwargs = {})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%erf, 1.0), kwargs = {})
# %mul_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%mul, %add), kwargs = {})
triton_poi_fused_add_div_erf_mul_0 = async_compile.triton('triton_poi_fused_add_div_erf_mul_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_div_erf_mul_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_div_erf_mul_0(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (x0), xmask)
tmp1 = 0.5
tmp2 = tmp0 * tmp1
tmp3 = 0.7071067811865475
tmp4 = tmp0 * tmp3
tmp5 = libdevice.erf(tmp4)
tmp6 = 1.0
tmp7 = tmp5 + tmp6
tmp8 = tmp2 * tmp7
tl.store(out_ptr0 + (x0), tmp8, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (4, 4), (4, 1))
assert_size_stride(primals_2, (4, ), (1, ))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [hidden_states], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_2, reinterpret_tensor(primals_3, (64, 4), (4, 1), 0), reinterpret_tensor(primals_1, (4, 4), (1, 4), 0), alpha=1, beta=1, out=buf0)
del primals_1
del primals_2
buf1 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [mul, truediv, erf, add, hidden_states_1], Original ATen: [aten.mul, aten.div, aten.erf, aten.add]
stream0 = get_raw_stream(0)
triton_poi_fused_add_div_erf_mul_0.run(buf0, buf1, 256, grid=grid(256), stream=stream0)
return (buf1, reinterpret_tensor(primals_3, (64, 4), (4, 1), 0), buf0, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| from _paritybench_helpers import _mock_config
import math
import torch
import torch.nn as nn
def gelu(x):
"""Implementation of the gelu activation function.
For information: OpenAI GPT's gelu is slightly different (and gives slightly different results):
0.5 * x * (1 + torch.tanh(math.sqrt(2 / math.pi) * (x + 0.044715 * torch.pow(x, 3))))
"""
return x * 0.5 * (1.0 + torch.erf(x / math.sqrt(2.0)))
class BERTIntermediate(nn.Module):
def __init__(self, config):
super(BERTIntermediate, self).__init__()
self.dense = nn.Linear(config.hidden_size, config.intermediate_size)
self.config = config
self.intermediate_act_fn = gelu
def forward(self, hidden_states):
hidden_states = self.dense(hidden_states)
hidden_states = self.intermediate_act_fn(hidden_states)
return hidden_states
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'config': _mock_config(hidden_size=4, intermediate_size=4)}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime.triton_helpers import libdevice
import math
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_add_div_erf_mul_0(in_ptr0, out_ptr0, xnumel, XBLOCK:
tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + x0, xmask)
tmp1 = 0.5
tmp2 = tmp0 * tmp1
tmp3 = 0.7071067811865475
tmp4 = tmp0 * tmp3
tmp5 = libdevice.erf(tmp4)
tmp6 = 1.0
tmp7 = tmp5 + tmp6
tmp8 = tmp2 * tmp7
tl.store(out_ptr0 + x0, tmp8, xmask)
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (4, 4), (4, 1))
assert_size_stride(primals_2, (4,), (1,))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
extern_kernels.addmm(primals_2, reinterpret_tensor(primals_3, (64,
4), (4, 1), 0), reinterpret_tensor(primals_1, (4, 4), (1, 4), 0
), alpha=1, beta=1, out=buf0)
del primals_1
del primals_2
buf1 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_add_div_erf_mul_0[grid(256)](buf0, buf1, 256,
XBLOCK=128, num_warps=4, num_stages=1)
return buf1, reinterpret_tensor(primals_3, (64, 4), (4, 1), 0), buf0
def gelu(x):
"""Implementation of the gelu activation function.
For information: OpenAI GPT's gelu is slightly different (and gives slightly different results):
0.5 * x * (1 + torch.tanh(math.sqrt(2 / math.pi) * (x + 0.044715 * torch.pow(x, 3))))
"""
return x * 0.5 * (1.0 + torch.erf(x / math.sqrt(2.0)))
class BERTIntermediateNew(nn.Module):
def __init__(self, config):
super(BERTIntermediateNew, self).__init__()
self.dense = nn.Linear(config.hidden_size, config.intermediate_size)
self.config = config
self.intermediate_act_fn = gelu
def forward(self, input_0):
primals_1 = self.dense.weight
primals_2 = self.dense.bias
primals_3 = input_0
output = call([primals_1, primals_2, primals_3])
return output[0]
| DAQuestionAnswering/Bert-n-Pals | BERTIntermediate | false | 6,896 | [
"MIT"
] | 1 | d5a288b9ac62259e70c249635108ba3906e19f00 | https://github.com/DAQuestionAnswering/Bert-n-Pals/tree/d5a288b9ac62259e70c249635108ba3906e19f00 | from _paritybench_helpers import _mock_config
import math
import torch
import torch.nn as nn
def gelu(x):
"""Implementation of the gelu activation function.
For information: OpenAI GPT's gelu is slightly different (and gives slightly different results):
0.5 * x * (1 + torch.tanh(math.sqrt(2 / math.pi) * (x + 0.044715 * torch.pow(x, 3))))
"""
return x * 0.5 * (1.0 + torch.erf(x / math.sqrt(2.0)))
class Model(nn.Module):
def __init__(self, config):
super().__init__()
self.dense = nn.Linear(config.hidden_size, config.intermediate_size)
self.config = config
self.intermediate_act_fn = gelu
def forward(self, hidden_states):
hidden_states = self.dense(hidden_states)
hidden_states = self.intermediate_act_fn(hidden_states)
return hidden_states
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return []
|
Decoder5 | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/5d/c5dw65h6nafj4s44sgiahjrq6lb3zgwonovnkpx75jkkuxpl34xg.py
# Topologically Sorted Source Nodes: [out], Original ATen: [aten.reflection_pad2d]
# Source node to ATen node mapping:
# out => _unsafe_index, _unsafe_index_1
# Graph fragment:
# %_unsafe_index : [num_users=1] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%primals_1, [None, None, %sub_1, None]), kwargs = {})
# %_unsafe_index_1 : [num_users=2] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%_unsafe_index, [None, None, None, %sub_1]), kwargs = {})
triton_poi_fused_reflection_pad2d_0 = async_compile.triton('triton_poi_fused_reflection_pad2d_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[131072],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_reflection_pad2d_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_reflection_pad2d_0(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 73728
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = tl.full([XBLOCK], True, tl.int1)
x0 = xindex % 6
x1 = (xindex // 6) % 6
x2 = (xindex // 36)
x3 = xindex
tmp0 = tl.load(in_ptr0 + (15 + ((-1)*(tl_math.abs((-3) + (tl_math.abs((-1) + x0))))) + ((-4)*(tl_math.abs((-3) + (tl_math.abs((-1) + x1))))) + (16*x2)), None, eviction_policy='evict_last')
tl.store(out_ptr0 + (x3), tmp0, None)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/kk/ckkeehad7xvjmxduxmwzjdm4zu3f5inttmd6kj4pju55xhwrnoea.py
# Topologically Sorted Source Nodes: [out_3], Original ATen: [aten.arange, aten.add, aten.mul, aten._to_copy]
# Source node to ATen node mapping:
# out_3 => add, add_1, convert_element_type, convert_element_type_1, iota_2, mul, mul_1
# Graph fragment:
# %iota_2 : [num_users=1] = call_function[target=torch.ops.prims.iota.default](args = (8,), kwargs = {start: 0, step: 1, dtype: torch.int64, device: cuda:0, requires_grad: False})
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%iota_2, 1), kwargs = {})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul, 0), kwargs = {})
# %convert_element_type : [num_users=1] = call_function[target=torch.ops.prims.convert_element_type.default](args = (%add, torch.float32), kwargs = {})
# %add_1 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%convert_element_type, 0.0), kwargs = {})
# %mul_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%add_1, 0.5), kwargs = {})
# %convert_element_type_1 : [num_users=3] = call_function[target=torch.ops.prims.convert_element_type.default](args = (%mul_1, torch.int64), kwargs = {})
triton_poi_fused__to_copy_add_arange_mul_1 = async_compile.triton('triton_poi_fused__to_copy_add_arange_mul_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[8],
filename=__file__,
triton_meta={'signature': {0: '*i64', 1: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0,), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__to_copy_add_arange_mul_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 0, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__to_copy_add_arange_mul_1(out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 8
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = x0
tmp1 = tmp0.to(tl.float32)
tmp2 = 0.5
tmp3 = tmp1 * tmp2
tmp4 = tmp3.to(tl.int32)
tl.store(out_ptr0 + (x0), tmp4, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/js/cjsibzhf6ubyye7gfes254fjkkbrqi7tzoox57ltdxf6tnfl6hr4.py
# Topologically Sorted Source Nodes: [out_1, out_2, out_3, out_4], Original ATen: [aten.convolution, aten.relu, aten._unsafe_index, aten.reflection_pad2d]
# Source node to ATen node mapping:
# out_1 => convolution
# out_2 => relu
# out_3 => _unsafe_index_2
# out_4 => _unsafe_index_3, _unsafe_index_4
# Graph fragment:
# %convolution : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%_unsafe_index_1, %primals_2, %primals_3, [1, 1], [0, 0], [1, 1], False, [0, 0], 1), kwargs = {})
# %relu : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%convolution,), kwargs = {})
# %_unsafe_index_2 : [num_users=1] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%relu, [None, None, %unsqueeze, %convert_element_type_1]), kwargs = {})
# %_unsafe_index_3 : [num_users=1] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%_unsafe_index_2, [None, None, %sub_5, None]), kwargs = {})
# %_unsafe_index_4 : [num_users=2] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%_unsafe_index_3, [None, None, None, %sub_5]), kwargs = {})
triton_poi_fused__unsafe_index_convolution_reflection_pad2d_relu_2 = async_compile.triton('triton_poi_fused__unsafe_index_convolution_reflection_pad2d_relu_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[262144],
filename=__file__,
triton_meta={'signature': {0: '*i64', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__unsafe_index_convolution_reflection_pad2d_relu_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 3, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__unsafe_index_convolution_reflection_pad2d_relu_2(in_ptr0, in_ptr1, in_ptr2, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 204800
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = tl.full([XBLOCK], True, tl.int1)
x1 = (xindex // 10) % 10
x0 = xindex % 10
x4 = (xindex // 100)
x2 = (xindex // 100) % 512
x7 = xindex
tmp0 = tl.load(in_ptr0 + (7 + ((-1)*(tl_math.abs((-7) + (tl_math.abs((-1) + x1)))))), None, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (7 + ((-1)*(tl_math.abs((-7) + (tl_math.abs((-1) + x0)))))), None, eviction_policy='evict_last')
tmp10 = tl.load(in_ptr2 + (x2), None, eviction_policy='evict_last')
tmp1 = tl.full([XBLOCK], 4, tl.int32)
tmp2 = tmp0 + tmp1
tmp3 = tmp0 < 0
tmp4 = tl.where(tmp3, tmp2, tmp0)
tmp6 = tmp5 + tmp1
tmp7 = tmp5 < 0
tmp8 = tl.where(tmp7, tmp6, tmp5)
tmp9 = tl.load(in_ptr1 + (tmp8 + (4*tmp4) + (16*x4)), None, eviction_policy='evict_last')
tmp11 = tmp9 + tmp10
tmp12 = tl.full([1], 0, tl.int32)
tmp13 = triton_helpers.maximum(tmp12, tmp11)
tl.store(out_ptr0 + (x7), tmp13, None)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/e2/ce2k4lym2nqhf3w5dsvz6zgxc3jsam7d572srar4v3rqbuyr34g6.py
# Topologically Sorted Source Nodes: [out_5, out_6, out_7], Original ATen: [aten.convolution, aten.relu, aten.reflection_pad2d]
# Source node to ATen node mapping:
# out_5 => convolution_1
# out_6 => relu_1
# out_7 => _unsafe_index_5, _unsafe_index_6
# Graph fragment:
# %convolution_1 : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%_unsafe_index_4, %primals_4, %primals_5, [1, 1], [0, 0], [1, 1], False, [0, 0], 1), kwargs = {})
# %relu_1 : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%convolution_1,), kwargs = {})
# %_unsafe_index_5 : [num_users=1] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%relu_1, [None, None, %sub_5, None]), kwargs = {})
# %_unsafe_index_6 : [num_users=2] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%_unsafe_index_5, [None, None, None, %sub_5]), kwargs = {})
triton_poi_fused_convolution_reflection_pad2d_relu_3 = async_compile.triton('triton_poi_fused_convolution_reflection_pad2d_relu_3', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[262144],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_reflection_pad2d_relu_3', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_reflection_pad2d_relu_3(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 204800
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = tl.full([XBLOCK], True, tl.int1)
x0 = xindex % 10
x1 = (xindex // 10) % 10
x4 = (xindex // 100)
x2 = (xindex // 100) % 512
x5 = xindex
tmp0 = tl.load(in_ptr0 + (63 + ((-1)*(tl_math.abs((-7) + (tl_math.abs((-1) + x0))))) + ((-8)*(tl_math.abs((-7) + (tl_math.abs((-1) + x1))))) + (64*x4)), None, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (x2), None, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tl.store(out_ptr0 + (x5), tmp4, None)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/7n/c7nyyvchtptp5rbqvks7cavg63gajcswx5tpkvypdmc7ocd2zaaz.py
# Topologically Sorted Source Nodes: [out_16], Original ATen: [aten.arange, aten.add, aten.mul, aten._to_copy]
# Source node to ATen node mapping:
# out_16 => add_4, add_5, convert_element_type_4, convert_element_type_5, iota_12, mul_4, mul_5
# Graph fragment:
# %iota_12 : [num_users=1] = call_function[target=torch.ops.prims.iota.default](args = (16,), kwargs = {start: 0, step: 1, dtype: torch.int64, device: cuda:0, requires_grad: False})
# %mul_4 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%iota_12, 1), kwargs = {})
# %add_4 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul_4, 0), kwargs = {})
# %convert_element_type_4 : [num_users=1] = call_function[target=torch.ops.prims.convert_element_type.default](args = (%add_4, torch.float32), kwargs = {})
# %add_5 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%convert_element_type_4, 0.0), kwargs = {})
# %mul_5 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%add_5, 0.5), kwargs = {})
# %convert_element_type_5 : [num_users=3] = call_function[target=torch.ops.prims.convert_element_type.default](args = (%mul_5, torch.int64), kwargs = {})
triton_poi_fused__to_copy_add_arange_mul_4 = async_compile.triton('triton_poi_fused__to_copy_add_arange_mul_4', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*i64', 1: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__to_copy_add_arange_mul_4', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 0, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__to_copy_add_arange_mul_4(out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = x0
tmp1 = tmp0.to(tl.float32)
tmp2 = 0.5
tmp3 = tmp1 * tmp2
tmp4 = tmp3.to(tl.int32)
tl.store(out_ptr0 + (x0), tmp4, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/bs/cbspecvkmrhwtp6geje57vqvyhtdbzqxeh4bkirrauq3bdgepjtc.py
# Topologically Sorted Source Nodes: [out_14, out_15, out_16, out_17], Original ATen: [aten.convolution, aten.relu, aten._unsafe_index, aten.reflection_pad2d]
# Source node to ATen node mapping:
# out_14 => convolution_4
# out_15 => relu_4
# out_16 => _unsafe_index_11
# out_17 => _unsafe_index_12, _unsafe_index_13
# Graph fragment:
# %convolution_4 : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%_unsafe_index_10, %primals_10, %primals_11, [1, 1], [0, 0], [1, 1], False, [0, 0], 1), kwargs = {})
# %relu_4 : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%convolution_4,), kwargs = {})
# %_unsafe_index_11 : [num_users=1] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%relu_4, [None, None, %unsqueeze_1, %convert_element_type_5]), kwargs = {})
# %_unsafe_index_12 : [num_users=1] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%_unsafe_index_11, [None, None, %sub_21, None]), kwargs = {})
# %_unsafe_index_13 : [num_users=2] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%_unsafe_index_12, [None, None, None, %sub_21]), kwargs = {})
triton_poi_fused__unsafe_index_convolution_reflection_pad2d_relu_5 = async_compile.triton('triton_poi_fused__unsafe_index_convolution_reflection_pad2d_relu_5', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[524288],
filename=__file__,
triton_meta={'signature': {0: '*i64', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__unsafe_index_convolution_reflection_pad2d_relu_5', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 3, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__unsafe_index_convolution_reflection_pad2d_relu_5(in_ptr0, in_ptr1, in_ptr2, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 331776
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = tl.full([XBLOCK], True, tl.int1)
x1 = (xindex // 18) % 18
x0 = xindex % 18
x4 = (xindex // 324)
x2 = (xindex // 324) % 256
x7 = xindex
tmp0 = tl.load(in_ptr0 + (15 + ((-1)*(tl_math.abs((-15) + (tl_math.abs((-1) + x1)))))), None, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (15 + ((-1)*(tl_math.abs((-15) + (tl_math.abs((-1) + x0)))))), None, eviction_policy='evict_last')
tmp10 = tl.load(in_ptr2 + (x2), None, eviction_policy='evict_last')
tmp1 = tl.full([XBLOCK], 8, tl.int32)
tmp2 = tmp0 + tmp1
tmp3 = tmp0 < 0
tmp4 = tl.where(tmp3, tmp2, tmp0)
tmp6 = tmp5 + tmp1
tmp7 = tmp5 < 0
tmp8 = tl.where(tmp7, tmp6, tmp5)
tmp9 = tl.load(in_ptr1 + (tmp8 + (8*tmp4) + (64*x4)), None, eviction_policy='evict_last')
tmp11 = tmp9 + tmp10
tmp12 = tl.full([1], 0, tl.int32)
tmp13 = triton_helpers.maximum(tmp12, tmp11)
tl.store(out_ptr0 + (x7), tmp13, None)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/6j/c6jeyn2lxggjjpib4soswcyajwy6iz7gvwfskycudxoh6ks5elxb.py
# Topologically Sorted Source Nodes: [out_18, out_19, out_20], Original ATen: [aten.convolution, aten.relu, aten.reflection_pad2d]
# Source node to ATen node mapping:
# out_18 => convolution_5
# out_19 => relu_5
# out_20 => _unsafe_index_14, _unsafe_index_15
# Graph fragment:
# %convolution_5 : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%_unsafe_index_13, %primals_12, %primals_13, [1, 1], [0, 0], [1, 1], False, [0, 0], 1), kwargs = {})
# %relu_5 : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%convolution_5,), kwargs = {})
# %_unsafe_index_14 : [num_users=1] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%relu_5, [None, None, %sub_21, None]), kwargs = {})
# %_unsafe_index_15 : [num_users=2] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%_unsafe_index_14, [None, None, None, %sub_21]), kwargs = {})
triton_poi_fused_convolution_reflection_pad2d_relu_6 = async_compile.triton('triton_poi_fused_convolution_reflection_pad2d_relu_6', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[524288],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_reflection_pad2d_relu_6', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_reflection_pad2d_relu_6(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 331776
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = tl.full([XBLOCK], True, tl.int1)
x0 = xindex % 18
x1 = (xindex // 18) % 18
x4 = (xindex // 324)
x2 = (xindex // 324) % 256
x5 = xindex
tmp0 = tl.load(in_ptr0 + (255 + ((-1)*(tl_math.abs((-15) + (tl_math.abs((-1) + x0))))) + ((-16)*(tl_math.abs((-15) + (tl_math.abs((-1) + x1))))) + (256*x4)), None, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (x2), None, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tl.store(out_ptr0 + (x5), tmp4, None)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/uw/cuwhadke7ul24n6hzb5sal45fu4ro7spo5s7tl5x4bo5jl7gz66f.py
# Topologically Sorted Source Nodes: [out_29], Original ATen: [aten.arange, aten.add, aten.mul, aten._to_copy]
# Source node to ATen node mapping:
# out_29 => add_8, add_9, convert_element_type_8, convert_element_type_9, iota_22, mul_8, mul_9
# Graph fragment:
# %iota_22 : [num_users=1] = call_function[target=torch.ops.prims.iota.default](args = (32,), kwargs = {start: 0, step: 1, dtype: torch.int64, device: cuda:0, requires_grad: False})
# %mul_8 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%iota_22, 1), kwargs = {})
# %add_8 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul_8, 0), kwargs = {})
# %convert_element_type_8 : [num_users=1] = call_function[target=torch.ops.prims.convert_element_type.default](args = (%add_8, torch.float32), kwargs = {})
# %add_9 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%convert_element_type_8, 0.0), kwargs = {})
# %mul_9 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%add_9, 0.5), kwargs = {})
# %convert_element_type_9 : [num_users=3] = call_function[target=torch.ops.prims.convert_element_type.default](args = (%mul_9, torch.int64), kwargs = {})
triton_poi_fused__to_copy_add_arange_mul_7 = async_compile.triton('triton_poi_fused__to_copy_add_arange_mul_7', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[32],
filename=__file__,
triton_meta={'signature': {0: '*i64', 1: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__to_copy_add_arange_mul_7', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 0, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__to_copy_add_arange_mul_7(out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 32
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = x0
tmp1 = tmp0.to(tl.float32)
tmp2 = 0.5
tmp3 = tmp1 * tmp2
tmp4 = tmp3.to(tl.int32)
tl.store(out_ptr0 + (x0), tmp4, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/h3/ch3jm3bt3vnato3q323q7wlnky2jb7vevymckrrxpv7cesdxjfra.py
# Topologically Sorted Source Nodes: [out_27, out_28, out_29, out_30], Original ATen: [aten.convolution, aten.relu, aten._unsafe_index, aten.reflection_pad2d]
# Source node to ATen node mapping:
# out_27 => convolution_8
# out_28 => relu_8
# out_29 => _unsafe_index_20
# out_30 => _unsafe_index_21, _unsafe_index_22
# Graph fragment:
# %convolution_8 : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%_unsafe_index_19, %primals_18, %primals_19, [1, 1], [0, 0], [1, 1], False, [0, 0], 1), kwargs = {})
# %relu_8 : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%convolution_8,), kwargs = {})
# %_unsafe_index_20 : [num_users=1] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%relu_8, [None, None, %unsqueeze_2, %convert_element_type_9]), kwargs = {})
# %_unsafe_index_21 : [num_users=1] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%_unsafe_index_20, [None, None, %sub_37, None]), kwargs = {})
# %_unsafe_index_22 : [num_users=2] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%_unsafe_index_21, [None, None, None, %sub_37]), kwargs = {})
triton_poi_fused__unsafe_index_convolution_reflection_pad2d_relu_8 = async_compile.triton('triton_poi_fused__unsafe_index_convolution_reflection_pad2d_relu_8', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[1048576],
filename=__file__,
triton_meta={'signature': {0: '*i64', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__unsafe_index_convolution_reflection_pad2d_relu_8', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 3, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__unsafe_index_convolution_reflection_pad2d_relu_8(in_ptr0, in_ptr1, in_ptr2, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 591872
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = tl.full([XBLOCK], True, tl.int1)
x1 = (xindex // 34) % 34
x0 = xindex % 34
x4 = (xindex // 1156)
x2 = (xindex // 1156) % 128
x7 = xindex
tmp0 = tl.load(in_ptr0 + (31 + ((-1)*(tl_math.abs((-31) + (tl_math.abs((-1) + x1)))))), None, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (31 + ((-1)*(tl_math.abs((-31) + (tl_math.abs((-1) + x0)))))), None, eviction_policy='evict_last')
tmp10 = tl.load(in_ptr2 + (x2), None, eviction_policy='evict_last')
tmp1 = tl.full([XBLOCK], 16, tl.int32)
tmp2 = tmp0 + tmp1
tmp3 = tmp0 < 0
tmp4 = tl.where(tmp3, tmp2, tmp0)
tmp6 = tmp5 + tmp1
tmp7 = tmp5 < 0
tmp8 = tl.where(tmp7, tmp6, tmp5)
tmp9 = tl.load(in_ptr1 + (tmp8 + (16*tmp4) + (256*x4)), None, eviction_policy='evict_last')
tmp11 = tmp9 + tmp10
tmp12 = tl.full([1], 0, tl.int32)
tmp13 = triton_helpers.maximum(tmp12, tmp11)
tl.store(out_ptr0 + (x7), tmp13, None)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/rh/crhey53viabe4hsp6ad6i2nee7q7zyv4njq6zse35tn7cpygnnbv.py
# Topologically Sorted Source Nodes: [out_31, out_32, out_33], Original ATen: [aten.convolution, aten.relu, aten.reflection_pad2d]
# Source node to ATen node mapping:
# out_31 => convolution_9
# out_32 => relu_9
# out_33 => _unsafe_index_23, _unsafe_index_24
# Graph fragment:
# %convolution_9 : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%_unsafe_index_22, %primals_20, %primals_21, [1, 1], [0, 0], [1, 1], False, [0, 0], 1), kwargs = {})
# %relu_9 : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%convolution_9,), kwargs = {})
# %_unsafe_index_23 : [num_users=1] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%relu_9, [None, None, %sub_37, None]), kwargs = {})
# %_unsafe_index_24 : [num_users=2] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%_unsafe_index_23, [None, None, None, %sub_37]), kwargs = {})
triton_poi_fused_convolution_reflection_pad2d_relu_9 = async_compile.triton('triton_poi_fused_convolution_reflection_pad2d_relu_9', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[1048576],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_reflection_pad2d_relu_9', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_reflection_pad2d_relu_9(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 591872
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = tl.full([XBLOCK], True, tl.int1)
x0 = xindex % 34
x1 = (xindex // 34) % 34
x4 = (xindex // 1156)
x2 = (xindex // 1156) % 128
x5 = xindex
tmp0 = tl.load(in_ptr0 + (1023 + ((-1)*(tl_math.abs((-31) + (tl_math.abs((-1) + x0))))) + ((-32)*(tl_math.abs((-31) + (tl_math.abs((-1) + x1))))) + (1024*x4)), None, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (x2), None, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tl.store(out_ptr0 + (x5), tmp4, None)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/bu/cbuvyw4fe3xbsdcjmotlovg3kxgvi3ofzc3qlcpjtkkuyytb22ws.py
# Topologically Sorted Source Nodes: [out_36], Original ATen: [aten.arange]
# Source node to ATen node mapping:
# out_36 => iota_28
# Graph fragment:
# %iota_28 : [num_users=2] = call_function[target=torch.ops.prims.iota.default](args = (64,), kwargs = {start: 0, step: 1, dtype: torch.int64, device: cuda:0, requires_grad: False})
triton_poi_fused_arange_10 = async_compile.triton('triton_poi_fused_arange_10', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*i64', 1: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_arange_10', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 0, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_arange_10(out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = x0
tl.store(out_ptr0 + (x0), tmp0, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/fn/cfndjiv3tblgvnqtytlbgh4w5inf7gba2up5wfjsuaevqfszyxak.py
# Topologically Sorted Source Nodes: [out_36], Original ATen: [aten.arange, aten.add, aten.mul, aten._to_copy]
# Source node to ATen node mapping:
# out_36 => add_12, add_13, convert_element_type_12, convert_element_type_13, mul_12, mul_13
# Graph fragment:
# %mul_12 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%iota_28, 1), kwargs = {})
# %add_12 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul_12, 0), kwargs = {})
# %convert_element_type_12 : [num_users=1] = call_function[target=torch.ops.prims.convert_element_type.default](args = (%add_12, torch.float32), kwargs = {})
# %add_13 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%convert_element_type_12, 0.0), kwargs = {})
# %mul_13 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%add_13, 0.5), kwargs = {})
# %convert_element_type_13 : [num_users=3] = call_function[target=torch.ops.prims.convert_element_type.default](args = (%mul_13, torch.int64), kwargs = {})
triton_poi_fused__to_copy_add_arange_mul_11 = async_compile.triton('triton_poi_fused__to_copy_add_arange_mul_11', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*i64', 1: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__to_copy_add_arange_mul_11', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 0, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__to_copy_add_arange_mul_11(out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = x0
tmp1 = tmp0.to(tl.float32)
tmp2 = 0.5
tmp3 = tmp1 * tmp2
tmp4 = tmp3.to(tl.int32)
tl.store(out_ptr0 + (x0), tmp4, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/rm/crmkhdiqj6544wyexgor4rnkds7gglwjet2cygw3o64zawf6zh7h.py
# Topologically Sorted Source Nodes: [out_34, out_35, out_36, out_37], Original ATen: [aten.convolution, aten.relu, aten._unsafe_index, aten.reflection_pad2d]
# Source node to ATen node mapping:
# out_34 => convolution_10
# out_35 => relu_10
# out_36 => _unsafe_index_25
# out_37 => _unsafe_index_26, _unsafe_index_27
# Graph fragment:
# %convolution_10 : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%_unsafe_index_24, %primals_22, %primals_23, [1, 1], [0, 0], [1, 1], False, [0, 0], 1), kwargs = {})
# %relu_10 : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%convolution_10,), kwargs = {})
# %_unsafe_index_25 : [num_users=1] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%relu_10, [None, None, %unsqueeze_3, %convert_element_type_13]), kwargs = {})
# %_unsafe_index_26 : [num_users=1] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%_unsafe_index_25, [None, None, %sub_45, None]), kwargs = {})
# %_unsafe_index_27 : [num_users=2] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%_unsafe_index_26, [None, None, None, %sub_45]), kwargs = {})
triton_poi_fused__unsafe_index_convolution_reflection_pad2d_relu_12 = async_compile.triton('triton_poi_fused__unsafe_index_convolution_reflection_pad2d_relu_12', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[2097152],
filename=__file__,
triton_meta={'signature': {0: '*i64', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__unsafe_index_convolution_reflection_pad2d_relu_12', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 3, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__unsafe_index_convolution_reflection_pad2d_relu_12(in_ptr0, in_ptr1, in_ptr2, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 1115136
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = (xindex // 66) % 66
x0 = xindex % 66
x4 = (xindex // 4356)
x2 = (xindex // 4356) % 64
x7 = xindex
tmp0 = tl.load(in_ptr0 + (63 + ((-1)*(tl_math.abs((-63) + (tl_math.abs((-1) + x1)))))), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (63 + ((-1)*(tl_math.abs((-63) + (tl_math.abs((-1) + x0)))))), xmask, eviction_policy='evict_last')
tmp10 = tl.load(in_ptr2 + (x2), xmask, eviction_policy='evict_last')
tmp1 = tl.full([XBLOCK], 32, tl.int32)
tmp2 = tmp0 + tmp1
tmp3 = tmp0 < 0
tmp4 = tl.where(tmp3, tmp2, tmp0)
tmp6 = tmp5 + tmp1
tmp7 = tmp5 < 0
tmp8 = tl.where(tmp7, tmp6, tmp5)
tmp9 = tl.load(in_ptr1 + (tmp8 + (32*tmp4) + (1024*x4)), xmask, eviction_policy='evict_last')
tmp11 = tmp9 + tmp10
tmp12 = tl.full([1], 0, tl.int32)
tmp13 = triton_helpers.maximum(tmp12, tmp11)
tl.store(out_ptr0 + (x7), tmp13, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/3t/c3tabqt44sb4glrpzzppdivi6ri4ru3tms7yd2a263tjtky77sll.py
# Topologically Sorted Source Nodes: [out_38, out_39, out_40], Original ATen: [aten.convolution, aten.relu, aten.reflection_pad2d]
# Source node to ATen node mapping:
# out_38 => convolution_11
# out_39 => relu_11
# out_40 => _unsafe_index_28, _unsafe_index_29
# Graph fragment:
# %convolution_11 : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%_unsafe_index_27, %primals_24, %primals_25, [1, 1], [0, 0], [1, 1], False, [0, 0], 1), kwargs = {})
# %relu_11 : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%convolution_11,), kwargs = {})
# %_unsafe_index_28 : [num_users=1] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%relu_11, [None, None, %sub_45, None]), kwargs = {})
# %_unsafe_index_29 : [num_users=2] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%_unsafe_index_28, [None, None, None, %sub_45]), kwargs = {})
triton_poi_fused_convolution_reflection_pad2d_relu_13 = async_compile.triton('triton_poi_fused_convolution_reflection_pad2d_relu_13', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[2097152],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_reflection_pad2d_relu_13', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_reflection_pad2d_relu_13(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 1115136
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 66
x1 = (xindex // 66) % 66
x4 = (xindex // 4356)
x2 = (xindex // 4356) % 64
x5 = xindex
tmp0 = tl.load(in_ptr0 + (4095 + ((-1)*(tl_math.abs((-63) + (tl_math.abs((-1) + x0))))) + ((-64)*(tl_math.abs((-63) + (tl_math.abs((-1) + x1))))) + (4096*x4)), xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (x2), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tl.store(out_ptr0 + (x5), tmp4, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/ln/cln46y6bcljj5xvgyl7h2nw5krdqozyeit2hebhehl5zbzlutyda.py
# Topologically Sorted Source Nodes: [out_41], Original ATen: [aten.convolution]
# Source node to ATen node mapping:
# out_41 => convolution_12
# Graph fragment:
# %convolution_12 : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%_unsafe_index_29, %primals_26, %primals_27, [1, 1], [0, 0], [1, 1], False, [0, 0], 1), kwargs = {})
triton_poi_fused_convolution_14 = async_compile.triton('triton_poi_fused_convolution_14', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[65536],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_14', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_14(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 49152
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = tl.full([XBLOCK], True, tl.int1)
x3 = xindex
x1 = (xindex // 4096) % 3
tmp0 = tl.load(in_out_ptr0 + (x3), None)
tmp1 = tl.load(in_ptr0 + (x1), None, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + (x3), tmp2, None)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/cj/ccjccu7zc7ogdpljitls2hefhpl6qq2kpfnn26koefpxdfdv7teq.py
# Topologically Sorted Source Nodes: [out_38, out_39], Original ATen: [aten.convolution, aten.relu, aten.threshold_backward]
# Source node to ATen node mapping:
# out_38 => convolution_11
# out_39 => relu_11
# Graph fragment:
# %convolution_11 : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%_unsafe_index_27, %primals_24, %primals_25, [1, 1], [0, 0], [1, 1], False, [0, 0], 1), kwargs = {})
# %relu_11 : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%convolution_11,), kwargs = {})
# %le_18 : [num_users=1] = call_function[target=torch.ops.aten.le.Scalar](args = (%relu_11, 0), kwargs = {})
triton_poi_fused_convolution_relu_threshold_backward_15 = async_compile.triton('triton_poi_fused_convolution_relu_threshold_backward_15', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[1048576],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*i1', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_relu_threshold_backward_15', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_relu_threshold_backward_15(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 1048576
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = tl.full([XBLOCK], True, tl.int1)
x3 = xindex
x1 = (xindex // 4096) % 64
tmp0 = tl.load(in_ptr0 + (x3), None)
tmp1 = tl.load(in_ptr1 + (x1), None, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(out_ptr0 + (x3), tmp6, None)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/at/catvmzd72emtja24agoetxsfbfoouankui4el2vao7ua26an3h5e.py
# Topologically Sorted Source Nodes: [out_34, out_35], Original ATen: [aten.convolution, aten.relu, aten.threshold_backward]
# Source node to ATen node mapping:
# out_34 => convolution_10
# out_35 => relu_10
# Graph fragment:
# %convolution_10 : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%_unsafe_index_24, %primals_22, %primals_23, [1, 1], [0, 0], [1, 1], False, [0, 0], 1), kwargs = {})
# %relu_10 : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%convolution_10,), kwargs = {})
# %le_37 : [num_users=1] = call_function[target=torch.ops.aten.le.Scalar](args = (%relu_10, 0), kwargs = {})
triton_poi_fused_convolution_relu_threshold_backward_16 = async_compile.triton('triton_poi_fused_convolution_relu_threshold_backward_16', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[262144],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*i1', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_relu_threshold_backward_16', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_relu_threshold_backward_16(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 262144
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = tl.full([XBLOCK], True, tl.int1)
x3 = xindex
x1 = (xindex // 1024) % 64
tmp0 = tl.load(in_ptr0 + (x3), None)
tmp1 = tl.load(in_ptr1 + (x1), None, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(out_ptr0 + (x3), tmp6, None)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/eu/ceu4vjwnu7t72daazfpt2wr6v3l3op2tzwyfbbm2wcynm5vjzmkk.py
# Topologically Sorted Source Nodes: [out_31, out_32], Original ATen: [aten.convolution, aten.relu, aten.threshold_backward]
# Source node to ATen node mapping:
# out_31 => convolution_9
# out_32 => relu_9
# Graph fragment:
# %convolution_9 : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%_unsafe_index_22, %primals_20, %primals_21, [1, 1], [0, 0], [1, 1], False, [0, 0], 1), kwargs = {})
# %relu_9 : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%convolution_9,), kwargs = {})
# %le_56 : [num_users=1] = call_function[target=torch.ops.aten.le.Scalar](args = (%relu_9, 0), kwargs = {})
triton_poi_fused_convolution_relu_threshold_backward_17 = async_compile.triton('triton_poi_fused_convolution_relu_threshold_backward_17', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[524288],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*i1', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_relu_threshold_backward_17', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_relu_threshold_backward_17(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 524288
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = tl.full([XBLOCK], True, tl.int1)
x3 = xindex
x1 = (xindex // 1024) % 128
tmp0 = tl.load(in_ptr0 + (x3), None)
tmp1 = tl.load(in_ptr1 + (x1), None, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(out_ptr0 + (x3), tmp6, None)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/g7/cg7u5oo3boouswvtbbyzyk3b6wbhd74gf7zcz66gyyh2akxdphhi.py
# Topologically Sorted Source Nodes: [out_27, out_28], Original ATen: [aten.convolution, aten.relu, aten.threshold_backward]
# Source node to ATen node mapping:
# out_27 => convolution_8
# out_28 => relu_8
# Graph fragment:
# %convolution_8 : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%_unsafe_index_19, %primals_18, %primals_19, [1, 1], [0, 0], [1, 1], False, [0, 0], 1), kwargs = {})
# %relu_8 : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%convolution_8,), kwargs = {})
# %le_75 : [num_users=1] = call_function[target=torch.ops.aten.le.Scalar](args = (%relu_8, 0), kwargs = {})
triton_poi_fused_convolution_relu_threshold_backward_18 = async_compile.triton('triton_poi_fused_convolution_relu_threshold_backward_18', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[131072],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*i1', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_relu_threshold_backward_18', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_relu_threshold_backward_18(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 131072
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = tl.full([XBLOCK], True, tl.int1)
x3 = xindex
x1 = (xindex // 256) % 128
tmp0 = tl.load(in_ptr0 + (x3), None)
tmp1 = tl.load(in_ptr1 + (x1), None, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(out_ptr0 + (x3), tmp6, None)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/cl/ccl6rdg5ilfmge5dt7ngbjb3ox5h4bpvbi7ffnd7mckl4lruweqh.py
# Topologically Sorted Source Nodes: [out_24, out_25], Original ATen: [aten.convolution, aten.relu, aten.threshold_backward]
# Source node to ATen node mapping:
# out_24 => convolution_7
# out_25 => relu_7
# Graph fragment:
# %convolution_7 : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%_unsafe_index_17, %primals_16, %primals_17, [1, 1], [0, 0], [1, 1], False, [0, 0], 1), kwargs = {})
# %relu_7 : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%convolution_7,), kwargs = {})
# %le_94 : [num_users=1] = call_function[target=torch.ops.aten.le.Scalar](args = (%relu_7, 0), kwargs = {})
triton_poi_fused_convolution_relu_threshold_backward_19 = async_compile.triton('triton_poi_fused_convolution_relu_threshold_backward_19', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[262144],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*i1', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_relu_threshold_backward_19', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_relu_threshold_backward_19(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 262144
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = tl.full([XBLOCK], True, tl.int1)
x3 = xindex
x1 = (xindex // 256) % 256
tmp0 = tl.load(in_ptr0 + (x3), None)
tmp1 = tl.load(in_ptr1 + (x1), None, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(out_ptr0 + (x3), tmp6, None)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/m3/cm3chxnerwp2olw2trs5cfxyfg5jjej3imjwwln6dvqacawhinod.py
# Topologically Sorted Source Nodes: [out_14, out_15], Original ATen: [aten.convolution, aten.relu, aten.threshold_backward]
# Source node to ATen node mapping:
# out_14 => convolution_4
# out_15 => relu_4
# Graph fragment:
# %convolution_4 : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%_unsafe_index_10, %primals_10, %primals_11, [1, 1], [0, 0], [1, 1], False, [0, 0], 1), kwargs = {})
# %relu_4 : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%convolution_4,), kwargs = {})
# %le_151 : [num_users=1] = call_function[target=torch.ops.aten.le.Scalar](args = (%relu_4, 0), kwargs = {})
triton_poi_fused_convolution_relu_threshold_backward_20 = async_compile.triton('triton_poi_fused_convolution_relu_threshold_backward_20', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[65536],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*i1', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_relu_threshold_backward_20', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_relu_threshold_backward_20(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 65536
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = tl.full([XBLOCK], True, tl.int1)
x3 = xindex
x1 = (xindex // 64) % 256
tmp0 = tl.load(in_ptr0 + (x3), None)
tmp1 = tl.load(in_ptr1 + (x1), None, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(out_ptr0 + (x3), tmp6, None)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/ts/ctswyyscojlcdj3bvsxhx5ujnib7w6opid7m4lucqvluzcl3n5pv.py
# Topologically Sorted Source Nodes: [out_11, out_12], Original ATen: [aten.convolution, aten.relu, aten.threshold_backward]
# Source node to ATen node mapping:
# out_11 => convolution_3
# out_12 => relu_3
# Graph fragment:
# %convolution_3 : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%_unsafe_index_8, %primals_8, %primals_9, [1, 1], [0, 0], [1, 1], False, [0, 0], 1), kwargs = {})
# %relu_3 : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%convolution_3,), kwargs = {})
# %le_170 : [num_users=1] = call_function[target=torch.ops.aten.le.Scalar](args = (%relu_3, 0), kwargs = {})
triton_poi_fused_convolution_relu_threshold_backward_21 = async_compile.triton('triton_poi_fused_convolution_relu_threshold_backward_21', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[131072],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*i1', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_relu_threshold_backward_21', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_relu_threshold_backward_21(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 131072
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = tl.full([XBLOCK], True, tl.int1)
x3 = xindex
x1 = (xindex // 64) % 512
tmp0 = tl.load(in_ptr0 + (x3), None)
tmp1 = tl.load(in_ptr1 + (x1), None, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(out_ptr0 + (x3), tmp6, None)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/bb/cbbl3sfaozoe4jbyrqweizaizeix4urgms7ffifscr2veiod4gcu.py
# Topologically Sorted Source Nodes: [out_1, out_2], Original ATen: [aten.convolution, aten.relu, aten.threshold_backward]
# Source node to ATen node mapping:
# out_1 => convolution
# out_2 => relu
# Graph fragment:
# %convolution : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%_unsafe_index_1, %primals_2, %primals_3, [1, 1], [0, 0], [1, 1], False, [0, 0], 1), kwargs = {})
# %relu : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%convolution,), kwargs = {})
# %le_227 : [num_users=1] = call_function[target=torch.ops.aten.le.Scalar](args = (%relu, 0), kwargs = {})
triton_poi_fused_convolution_relu_threshold_backward_22 = async_compile.triton('triton_poi_fused_convolution_relu_threshold_backward_22', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[32768],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*i1', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_relu_threshold_backward_22', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_relu_threshold_backward_22(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 32768
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = tl.full([XBLOCK], True, tl.int1)
x3 = xindex
x1 = (xindex // 16) % 512
tmp0 = tl.load(in_ptr0 + (x3), None)
tmp1 = tl.load(in_ptr1 + (x1), None, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(out_ptr0 + (x3), tmp6, None)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9, primals_10, primals_11, primals_12, primals_13, primals_14, primals_15, primals_16, primals_17, primals_18, primals_19, primals_20, primals_21, primals_22, primals_23, primals_24, primals_25, primals_26, primals_27 = args
args.clear()
assert_size_stride(primals_1, (4, 512, 4, 4), (8192, 16, 4, 1))
assert_size_stride(primals_2, (512, 512, 3, 3), (4608, 9, 3, 1))
assert_size_stride(primals_3, (512, ), (1, ))
assert_size_stride(primals_4, (512, 512, 3, 3), (4608, 9, 3, 1))
assert_size_stride(primals_5, (512, ), (1, ))
assert_size_stride(primals_6, (512, 512, 3, 3), (4608, 9, 3, 1))
assert_size_stride(primals_7, (512, ), (1, ))
assert_size_stride(primals_8, (512, 512, 3, 3), (4608, 9, 3, 1))
assert_size_stride(primals_9, (512, ), (1, ))
assert_size_stride(primals_10, (256, 512, 3, 3), (4608, 9, 3, 1))
assert_size_stride(primals_11, (256, ), (1, ))
assert_size_stride(primals_12, (256, 256, 3, 3), (2304, 9, 3, 1))
assert_size_stride(primals_13, (256, ), (1, ))
assert_size_stride(primals_14, (256, 256, 3, 3), (2304, 9, 3, 1))
assert_size_stride(primals_15, (256, ), (1, ))
assert_size_stride(primals_16, (256, 256, 3, 3), (2304, 9, 3, 1))
assert_size_stride(primals_17, (256, ), (1, ))
assert_size_stride(primals_18, (128, 256, 3, 3), (2304, 9, 3, 1))
assert_size_stride(primals_19, (128, ), (1, ))
assert_size_stride(primals_20, (128, 128, 3, 3), (1152, 9, 3, 1))
assert_size_stride(primals_21, (128, ), (1, ))
assert_size_stride(primals_22, (64, 128, 3, 3), (1152, 9, 3, 1))
assert_size_stride(primals_23, (64, ), (1, ))
assert_size_stride(primals_24, (64, 64, 3, 3), (576, 9, 3, 1))
assert_size_stride(primals_25, (64, ), (1, ))
assert_size_stride(primals_26, (3, 64, 3, 3), (576, 9, 3, 1))
assert_size_stride(primals_27, (3, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 512, 6, 6), (18432, 36, 6, 1), torch.float32)
# Topologically Sorted Source Nodes: [out], Original ATen: [aten.reflection_pad2d]
stream0 = get_raw_stream(0)
triton_poi_fused_reflection_pad2d_0.run(primals_1, buf0, 73728, grid=grid(73728), stream=stream0)
del primals_1
# Topologically Sorted Source Nodes: [out_1], Original ATen: [aten.convolution]
buf1 = extern_kernels.convolution(buf0, primals_2, stride=(1, 1), padding=(0, 0), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf1, (4, 512, 4, 4), (8192, 16, 4, 1))
buf2 = empty_strided_cuda((8, ), (1, ), torch.int64)
# Topologically Sorted Source Nodes: [out_3], Original ATen: [aten.arange, aten.add, aten.mul, aten._to_copy]
triton_poi_fused__to_copy_add_arange_mul_1.run(buf2, 8, grid=grid(8), stream=stream0)
buf3 = empty_strided_cuda((4, 512, 10, 10), (51200, 100, 10, 1), torch.float32)
# Topologically Sorted Source Nodes: [out_1, out_2, out_3, out_4], Original ATen: [aten.convolution, aten.relu, aten._unsafe_index, aten.reflection_pad2d]
triton_poi_fused__unsafe_index_convolution_reflection_pad2d_relu_2.run(buf2, buf1, primals_3, buf3, 204800, grid=grid(204800), stream=stream0)
# Topologically Sorted Source Nodes: [out_5], Original ATen: [aten.convolution]
buf4 = extern_kernels.convolution(buf3, primals_4, stride=(1, 1), padding=(0, 0), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf4, (4, 512, 8, 8), (32768, 64, 8, 1))
buf5 = empty_strided_cuda((4, 512, 10, 10), (51200, 100, 10, 1), torch.float32)
# Topologically Sorted Source Nodes: [out_5, out_6, out_7], Original ATen: [aten.convolution, aten.relu, aten.reflection_pad2d]
triton_poi_fused_convolution_reflection_pad2d_relu_3.run(buf4, primals_5, buf5, 204800, grid=grid(204800), stream=stream0)
# Topologically Sorted Source Nodes: [out_8], Original ATen: [aten.convolution]
buf6 = extern_kernels.convolution(buf5, primals_6, stride=(1, 1), padding=(0, 0), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf6, (4, 512, 8, 8), (32768, 64, 8, 1))
buf7 = empty_strided_cuda((4, 512, 10, 10), (51200, 100, 10, 1), torch.float32)
# Topologically Sorted Source Nodes: [out_8, out_9, out_10], Original ATen: [aten.convolution, aten.relu, aten.reflection_pad2d]
triton_poi_fused_convolution_reflection_pad2d_relu_3.run(buf6, primals_7, buf7, 204800, grid=grid(204800), stream=stream0)
# Topologically Sorted Source Nodes: [out_11], Original ATen: [aten.convolution]
buf8 = extern_kernels.convolution(buf7, primals_8, stride=(1, 1), padding=(0, 0), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf8, (4, 512, 8, 8), (32768, 64, 8, 1))
buf9 = empty_strided_cuda((4, 512, 10, 10), (51200, 100, 10, 1), torch.float32)
# Topologically Sorted Source Nodes: [out_11, out_12, out_13], Original ATen: [aten.convolution, aten.relu, aten.reflection_pad2d]
triton_poi_fused_convolution_reflection_pad2d_relu_3.run(buf8, primals_9, buf9, 204800, grid=grid(204800), stream=stream0)
# Topologically Sorted Source Nodes: [out_14], Original ATen: [aten.convolution]
buf10 = extern_kernels.convolution(buf9, primals_10, stride=(1, 1), padding=(0, 0), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf10, (4, 256, 8, 8), (16384, 64, 8, 1))
buf11 = empty_strided_cuda((16, ), (1, ), torch.int64)
# Topologically Sorted Source Nodes: [out_16], Original ATen: [aten.arange, aten.add, aten.mul, aten._to_copy]
triton_poi_fused__to_copy_add_arange_mul_4.run(buf11, 16, grid=grid(16), stream=stream0)
buf12 = empty_strided_cuda((4, 256, 18, 18), (82944, 324, 18, 1), torch.float32)
# Topologically Sorted Source Nodes: [out_14, out_15, out_16, out_17], Original ATen: [aten.convolution, aten.relu, aten._unsafe_index, aten.reflection_pad2d]
triton_poi_fused__unsafe_index_convolution_reflection_pad2d_relu_5.run(buf11, buf10, primals_11, buf12, 331776, grid=grid(331776), stream=stream0)
# Topologically Sorted Source Nodes: [out_18], Original ATen: [aten.convolution]
buf13 = extern_kernels.convolution(buf12, primals_12, stride=(1, 1), padding=(0, 0), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf13, (4, 256, 16, 16), (65536, 256, 16, 1))
buf14 = empty_strided_cuda((4, 256, 18, 18), (82944, 324, 18, 1), torch.float32)
# Topologically Sorted Source Nodes: [out_18, out_19, out_20], Original ATen: [aten.convolution, aten.relu, aten.reflection_pad2d]
triton_poi_fused_convolution_reflection_pad2d_relu_6.run(buf13, primals_13, buf14, 331776, grid=grid(331776), stream=stream0)
# Topologically Sorted Source Nodes: [out_21], Original ATen: [aten.convolution]
buf15 = extern_kernels.convolution(buf14, primals_14, stride=(1, 1), padding=(0, 0), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf15, (4, 256, 16, 16), (65536, 256, 16, 1))
buf16 = empty_strided_cuda((4, 256, 18, 18), (82944, 324, 18, 1), torch.float32)
# Topologically Sorted Source Nodes: [out_21, out_22, out_23], Original ATen: [aten.convolution, aten.relu, aten.reflection_pad2d]
triton_poi_fused_convolution_reflection_pad2d_relu_6.run(buf15, primals_15, buf16, 331776, grid=grid(331776), stream=stream0)
# Topologically Sorted Source Nodes: [out_24], Original ATen: [aten.convolution]
buf17 = extern_kernels.convolution(buf16, primals_16, stride=(1, 1), padding=(0, 0), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf17, (4, 256, 16, 16), (65536, 256, 16, 1))
buf18 = empty_strided_cuda((4, 256, 18, 18), (82944, 324, 18, 1), torch.float32)
# Topologically Sorted Source Nodes: [out_24, out_25, out_26], Original ATen: [aten.convolution, aten.relu, aten.reflection_pad2d]
triton_poi_fused_convolution_reflection_pad2d_relu_6.run(buf17, primals_17, buf18, 331776, grid=grid(331776), stream=stream0)
# Topologically Sorted Source Nodes: [out_27], Original ATen: [aten.convolution]
buf19 = extern_kernels.convolution(buf18, primals_18, stride=(1, 1), padding=(0, 0), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf19, (4, 128, 16, 16), (32768, 256, 16, 1))
buf20 = empty_strided_cuda((32, ), (1, ), torch.int64)
# Topologically Sorted Source Nodes: [out_29], Original ATen: [aten.arange, aten.add, aten.mul, aten._to_copy]
triton_poi_fused__to_copy_add_arange_mul_7.run(buf20, 32, grid=grid(32), stream=stream0)
buf21 = empty_strided_cuda((4, 128, 34, 34), (147968, 1156, 34, 1), torch.float32)
# Topologically Sorted Source Nodes: [out_27, out_28, out_29, out_30], Original ATen: [aten.convolution, aten.relu, aten._unsafe_index, aten.reflection_pad2d]
triton_poi_fused__unsafe_index_convolution_reflection_pad2d_relu_8.run(buf20, buf19, primals_19, buf21, 591872, grid=grid(591872), stream=stream0)
# Topologically Sorted Source Nodes: [out_31], Original ATen: [aten.convolution]
buf22 = extern_kernels.convolution(buf21, primals_20, stride=(1, 1), padding=(0, 0), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf22, (4, 128, 32, 32), (131072, 1024, 32, 1))
buf23 = empty_strided_cuda((4, 128, 34, 34), (147968, 1156, 34, 1), torch.float32)
# Topologically Sorted Source Nodes: [out_31, out_32, out_33], Original ATen: [aten.convolution, aten.relu, aten.reflection_pad2d]
triton_poi_fused_convolution_reflection_pad2d_relu_9.run(buf22, primals_21, buf23, 591872, grid=grid(591872), stream=stream0)
# Topologically Sorted Source Nodes: [out_34], Original ATen: [aten.convolution]
buf24 = extern_kernels.convolution(buf23, primals_22, stride=(1, 1), padding=(0, 0), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf24, (4, 64, 32, 32), (65536, 1024, 32, 1))
buf25 = empty_strided_cuda((64, ), (1, ), torch.int64)
# Topologically Sorted Source Nodes: [out_36], Original ATen: [aten.arange]
triton_poi_fused_arange_10.run(buf25, 64, grid=grid(64), stream=stream0)
buf26 = empty_strided_cuda((64, ), (1, ), torch.int64)
# Topologically Sorted Source Nodes: [out_36], Original ATen: [aten.arange, aten.add, aten.mul, aten._to_copy]
triton_poi_fused__to_copy_add_arange_mul_11.run(buf26, 64, grid=grid(64), stream=stream0)
buf27 = empty_strided_cuda((4, 64, 66, 66), (278784, 4356, 66, 1), torch.float32)
# Topologically Sorted Source Nodes: [out_34, out_35, out_36, out_37], Original ATen: [aten.convolution, aten.relu, aten._unsafe_index, aten.reflection_pad2d]
triton_poi_fused__unsafe_index_convolution_reflection_pad2d_relu_12.run(buf26, buf24, primals_23, buf27, 1115136, grid=grid(1115136), stream=stream0)
# Topologically Sorted Source Nodes: [out_38], Original ATen: [aten.convolution]
buf28 = extern_kernels.convolution(buf27, primals_24, stride=(1, 1), padding=(0, 0), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf28, (4, 64, 64, 64), (262144, 4096, 64, 1))
buf29 = empty_strided_cuda((4, 64, 66, 66), (278784, 4356, 66, 1), torch.float32)
# Topologically Sorted Source Nodes: [out_38, out_39, out_40], Original ATen: [aten.convolution, aten.relu, aten.reflection_pad2d]
triton_poi_fused_convolution_reflection_pad2d_relu_13.run(buf28, primals_25, buf29, 1115136, grid=grid(1115136), stream=stream0)
# Topologically Sorted Source Nodes: [out_41], Original ATen: [aten.convolution]
buf30 = extern_kernels.convolution(buf29, primals_26, stride=(1, 1), padding=(0, 0), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf30, (4, 3, 64, 64), (12288, 4096, 64, 1))
buf31 = buf30; del buf30 # reuse
# Topologically Sorted Source Nodes: [out_41], Original ATen: [aten.convolution]
triton_poi_fused_convolution_14.run(buf31, primals_27, 49152, grid=grid(49152), stream=stream0)
del primals_27
buf32 = empty_strided_cuda((4, 64, 64, 64), (262144, 4096, 64, 1), torch.bool)
# Topologically Sorted Source Nodes: [out_38, out_39], Original ATen: [aten.convolution, aten.relu, aten.threshold_backward]
triton_poi_fused_convolution_relu_threshold_backward_15.run(buf28, primals_25, buf32, 1048576, grid=grid(1048576), stream=stream0)
del buf28
del primals_25
buf33 = empty_strided_cuda((4, 64, 32, 32), (65536, 1024, 32, 1), torch.bool)
# Topologically Sorted Source Nodes: [out_34, out_35], Original ATen: [aten.convolution, aten.relu, aten.threshold_backward]
triton_poi_fused_convolution_relu_threshold_backward_16.run(buf24, primals_23, buf33, 262144, grid=grid(262144), stream=stream0)
del buf24
del primals_23
buf34 = empty_strided_cuda((4, 128, 32, 32), (131072, 1024, 32, 1), torch.bool)
# Topologically Sorted Source Nodes: [out_31, out_32], Original ATen: [aten.convolution, aten.relu, aten.threshold_backward]
triton_poi_fused_convolution_relu_threshold_backward_17.run(buf22, primals_21, buf34, 524288, grid=grid(524288), stream=stream0)
del buf22
del primals_21
buf35 = empty_strided_cuda((4, 128, 16, 16), (32768, 256, 16, 1), torch.bool)
# Topologically Sorted Source Nodes: [out_27, out_28], Original ATen: [aten.convolution, aten.relu, aten.threshold_backward]
triton_poi_fused_convolution_relu_threshold_backward_18.run(buf19, primals_19, buf35, 131072, grid=grid(131072), stream=stream0)
del buf19
del primals_19
buf36 = empty_strided_cuda((4, 256, 16, 16), (65536, 256, 16, 1), torch.bool)
# Topologically Sorted Source Nodes: [out_24, out_25], Original ATen: [aten.convolution, aten.relu, aten.threshold_backward]
triton_poi_fused_convolution_relu_threshold_backward_19.run(buf17, primals_17, buf36, 262144, grid=grid(262144), stream=stream0)
del buf17
del primals_17
buf37 = empty_strided_cuda((4, 256, 16, 16), (65536, 256, 16, 1), torch.bool)
# Topologically Sorted Source Nodes: [out_21, out_22], Original ATen: [aten.convolution, aten.relu, aten.threshold_backward]
triton_poi_fused_convolution_relu_threshold_backward_19.run(buf15, primals_15, buf37, 262144, grid=grid(262144), stream=stream0)
del buf15
del primals_15
buf38 = empty_strided_cuda((4, 256, 16, 16), (65536, 256, 16, 1), torch.bool)
# Topologically Sorted Source Nodes: [out_18, out_19], Original ATen: [aten.convolution, aten.relu, aten.threshold_backward]
triton_poi_fused_convolution_relu_threshold_backward_19.run(buf13, primals_13, buf38, 262144, grid=grid(262144), stream=stream0)
del buf13
del primals_13
buf39 = empty_strided_cuda((4, 256, 8, 8), (16384, 64, 8, 1), torch.bool)
# Topologically Sorted Source Nodes: [out_14, out_15], Original ATen: [aten.convolution, aten.relu, aten.threshold_backward]
triton_poi_fused_convolution_relu_threshold_backward_20.run(buf10, primals_11, buf39, 65536, grid=grid(65536), stream=stream0)
del buf10
del primals_11
buf40 = empty_strided_cuda((4, 512, 8, 8), (32768, 64, 8, 1), torch.bool)
# Topologically Sorted Source Nodes: [out_11, out_12], Original ATen: [aten.convolution, aten.relu, aten.threshold_backward]
triton_poi_fused_convolution_relu_threshold_backward_21.run(buf8, primals_9, buf40, 131072, grid=grid(131072), stream=stream0)
del buf8
del primals_9
buf41 = empty_strided_cuda((4, 512, 8, 8), (32768, 64, 8, 1), torch.bool)
# Topologically Sorted Source Nodes: [out_8, out_9], Original ATen: [aten.convolution, aten.relu, aten.threshold_backward]
triton_poi_fused_convolution_relu_threshold_backward_21.run(buf6, primals_7, buf41, 131072, grid=grid(131072), stream=stream0)
del buf6
del primals_7
buf42 = empty_strided_cuda((4, 512, 8, 8), (32768, 64, 8, 1), torch.bool)
# Topologically Sorted Source Nodes: [out_5, out_6], Original ATen: [aten.convolution, aten.relu, aten.threshold_backward]
triton_poi_fused_convolution_relu_threshold_backward_21.run(buf4, primals_5, buf42, 131072, grid=grid(131072), stream=stream0)
del buf4
del primals_5
buf43 = empty_strided_cuda((4, 512, 4, 4), (8192, 16, 4, 1), torch.bool)
# Topologically Sorted Source Nodes: [out_1, out_2], Original ATen: [aten.convolution, aten.relu, aten.threshold_backward]
triton_poi_fused_convolution_relu_threshold_backward_22.run(buf1, primals_3, buf43, 32768, grid=grid(32768), stream=stream0)
del buf1
del primals_3
return (buf31, primals_2, primals_4, primals_6, primals_8, primals_10, primals_12, primals_14, primals_16, primals_18, primals_20, primals_22, primals_24, primals_26, buf0, buf2, buf3, buf5, buf7, buf9, buf11, buf12, buf14, buf16, buf18, buf20, buf21, buf23, buf25, buf26, buf27, buf29, buf32, buf33, buf34, buf35, buf36, buf37, buf38, buf39, buf40, buf41, buf42, buf43, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 512, 4, 4), (8192, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((512, 512, 3, 3), (4608, 9, 3, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((512, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((512, 512, 3, 3), (4608, 9, 3, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((512, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((512, 512, 3, 3), (4608, 9, 3, 1), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((512, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_8 = rand_strided((512, 512, 3, 3), (4608, 9, 3, 1), device='cuda:0', dtype=torch.float32)
primals_9 = rand_strided((512, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_10 = rand_strided((256, 512, 3, 3), (4608, 9, 3, 1), device='cuda:0', dtype=torch.float32)
primals_11 = rand_strided((256, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_12 = rand_strided((256, 256, 3, 3), (2304, 9, 3, 1), device='cuda:0', dtype=torch.float32)
primals_13 = rand_strided((256, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_14 = rand_strided((256, 256, 3, 3), (2304, 9, 3, 1), device='cuda:0', dtype=torch.float32)
primals_15 = rand_strided((256, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_16 = rand_strided((256, 256, 3, 3), (2304, 9, 3, 1), device='cuda:0', dtype=torch.float32)
primals_17 = rand_strided((256, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_18 = rand_strided((128, 256, 3, 3), (2304, 9, 3, 1), device='cuda:0', dtype=torch.float32)
primals_19 = rand_strided((128, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_20 = rand_strided((128, 128, 3, 3), (1152, 9, 3, 1), device='cuda:0', dtype=torch.float32)
primals_21 = rand_strided((128, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_22 = rand_strided((64, 128, 3, 3), (1152, 9, 3, 1), device='cuda:0', dtype=torch.float32)
primals_23 = rand_strided((64, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_24 = rand_strided((64, 64, 3, 3), (576, 9, 3, 1), device='cuda:0', dtype=torch.float32)
primals_25 = rand_strided((64, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_26 = rand_strided((3, 64, 3, 3), (576, 9, 3, 1), device='cuda:0', dtype=torch.float32)
primals_27 = rand_strided((3, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9, primals_10, primals_11, primals_12, primals_13, primals_14, primals_15, primals_16, primals_17, primals_18, primals_19, primals_20, primals_21, primals_22, primals_23, primals_24, primals_25, primals_26, primals_27])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
class Decoder5(nn.Module):
def __init__(self):
super(Decoder5, self).__init__()
self.reflecPad15 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv15 = nn.Conv2d(512, 512, 3, 1, 0)
self.relu15 = nn.ReLU(inplace=True)
self.unpool = nn.UpsamplingNearest2d(scale_factor=2)
self.reflecPad16 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv16 = nn.Conv2d(512, 512, 3, 1, 0)
self.relu16 = nn.ReLU(inplace=True)
self.reflecPad17 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv17 = nn.Conv2d(512, 512, 3, 1, 0)
self.relu17 = nn.ReLU(inplace=True)
self.reflecPad18 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv18 = nn.Conv2d(512, 512, 3, 1, 0)
self.relu18 = nn.ReLU(inplace=True)
self.reflecPad19 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv19 = nn.Conv2d(512, 256, 3, 1, 0)
self.relu19 = nn.ReLU(inplace=True)
self.unpool2 = nn.UpsamplingNearest2d(scale_factor=2)
self.reflecPad20 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv20 = nn.Conv2d(256, 256, 3, 1, 0)
self.relu20 = nn.ReLU(inplace=True)
self.reflecPad21 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv21 = nn.Conv2d(256, 256, 3, 1, 0)
self.relu21 = nn.ReLU(inplace=True)
self.reflecPad22 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv22 = nn.Conv2d(256, 256, 3, 1, 0)
self.relu22 = nn.ReLU(inplace=True)
self.reflecPad23 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv23 = nn.Conv2d(256, 128, 3, 1, 0)
self.relu23 = nn.ReLU(inplace=True)
self.unpool3 = nn.UpsamplingNearest2d(scale_factor=2)
self.reflecPad24 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv24 = nn.Conv2d(128, 128, 3, 1, 0)
self.relu24 = nn.ReLU(inplace=True)
self.reflecPad25 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv25 = nn.Conv2d(128, 64, 3, 1, 0)
self.relu25 = nn.ReLU(inplace=True)
self.unpool4 = nn.UpsamplingNearest2d(scale_factor=2)
self.reflecPad26 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv26 = nn.Conv2d(64, 64, 3, 1, 0)
self.relu26 = nn.ReLU(inplace=True)
self.reflecPad27 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv27 = nn.Conv2d(64, 3, 3, 1, 0)
def forward(self, x):
out = self.reflecPad15(x)
out = self.conv15(out)
out = self.relu15(out)
out = self.unpool(out)
out = self.reflecPad16(out)
out = self.conv16(out)
out = self.relu16(out)
out = self.reflecPad17(out)
out = self.conv17(out)
out = self.relu17(out)
out = self.reflecPad18(out)
out = self.conv18(out)
out = self.relu18(out)
out = self.reflecPad19(out)
out = self.conv19(out)
out = self.relu19(out)
out = self.unpool2(out)
out = self.reflecPad20(out)
out = self.conv20(out)
out = self.relu20(out)
out = self.reflecPad21(out)
out = self.conv21(out)
out = self.relu21(out)
out = self.reflecPad22(out)
out = self.conv22(out)
out = self.relu22(out)
out = self.reflecPad23(out)
out = self.conv23(out)
out = self.relu23(out)
out = self.unpool3(out)
out = self.reflecPad24(out)
out = self.conv24(out)
out = self.relu24(out)
out = self.reflecPad25(out)
out = self.conv25(out)
out = self.relu25(out)
out = self.unpool4(out)
out = self.reflecPad26(out)
out = self.conv26(out)
out = self.relu26(out)
out = self.reflecPad27(out)
out = self.conv27(out)
return out
def get_inputs():
return [torch.rand([4, 512, 4, 4])]
def get_init_inputs():
return [[], {}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import math as tl_math
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_poi_fused_reflection_pad2d_0(in_ptr0, out_ptr0, xnumel, XBLOCK:
tl.constexpr):
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
tl.full([XBLOCK], True, tl.int1)
x0 = xindex % 6
x1 = xindex // 6 % 6
x2 = xindex // 36
x3 = xindex
tmp0 = tl.load(in_ptr0 + (15 + -1 * tl_math.abs(-3 + tl_math.abs(-1 +
x0)) + -4 * tl_math.abs(-3 + tl_math.abs(-1 + x1)) + 16 * x2), None,
eviction_policy='evict_last')
tl.store(out_ptr0 + x3, tmp0, None)
@triton.jit
def triton_poi_fused__to_copy_add_arange_mul_1(out_ptr0, xnumel, XBLOCK: tl
.constexpr):
xnumel = 8
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = x0
tmp1 = tmp0.to(tl.float32)
tmp2 = 0.5
tmp3 = tmp1 * tmp2
tmp4 = tmp3.to(tl.int32)
tl.store(out_ptr0 + x0, tmp4, xmask)
@triton.jit
def triton_poi_fused__unsafe_index_convolution_reflection_pad2d_relu_2(in_ptr0,
in_ptr1, in_ptr2, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
tl.full([XBLOCK], True, tl.int1)
x1 = xindex // 10 % 10
x0 = xindex % 10
x4 = xindex // 100
x2 = xindex // 100 % 512
x7 = xindex
tmp0 = tl.load(in_ptr0 + (7 + -1 * tl_math.abs(-7 + tl_math.abs(-1 + x1
))), None, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (7 + -1 * tl_math.abs(-7 + tl_math.abs(-1 + x0
))), None, eviction_policy='evict_last')
tmp10 = tl.load(in_ptr2 + x2, None, eviction_policy='evict_last')
tmp1 = tl.full([XBLOCK], 4, tl.int32)
tmp2 = tmp0 + tmp1
tmp3 = tmp0 < 0
tmp4 = tl.where(tmp3, tmp2, tmp0)
tmp6 = tmp5 + tmp1
tmp7 = tmp5 < 0
tmp8 = tl.where(tmp7, tmp6, tmp5)
tmp9 = tl.load(in_ptr1 + (tmp8 + 4 * tmp4 + 16 * x4), None,
eviction_policy='evict_last')
tmp11 = tmp9 + tmp10
tmp12 = tl.full([1], 0, tl.int32)
tmp13 = triton_helpers.maximum(tmp12, tmp11)
tl.store(out_ptr0 + x7, tmp13, None)
@triton.jit
def triton_poi_fused_convolution_reflection_pad2d_relu_3(in_ptr0, in_ptr1,
out_ptr0, xnumel, XBLOCK: tl.constexpr):
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
tl.full([XBLOCK], True, tl.int1)
x0 = xindex % 10
x1 = xindex // 10 % 10
x4 = xindex // 100
x2 = xindex // 100 % 512
x5 = xindex
tmp0 = tl.load(in_ptr0 + (63 + -1 * tl_math.abs(-7 + tl_math.abs(-1 +
x0)) + -8 * tl_math.abs(-7 + tl_math.abs(-1 + x1)) + 64 * x4), None,
eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + x2, None, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tl.store(out_ptr0 + x5, tmp4, None)
@triton.jit
def triton_poi_fused__to_copy_add_arange_mul_4(out_ptr0, xnumel, XBLOCK: tl
.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = x0
tmp1 = tmp0.to(tl.float32)
tmp2 = 0.5
tmp3 = tmp1 * tmp2
tmp4 = tmp3.to(tl.int32)
tl.store(out_ptr0 + x0, tmp4, xmask)
@triton.jit
def triton_poi_fused__unsafe_index_convolution_reflection_pad2d_relu_5(in_ptr0,
in_ptr1, in_ptr2, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
tl.full([XBLOCK], True, tl.int1)
x1 = xindex // 18 % 18
x0 = xindex % 18
x4 = xindex // 324
x2 = xindex // 324 % 256
x7 = xindex
tmp0 = tl.load(in_ptr0 + (15 + -1 * tl_math.abs(-15 + tl_math.abs(-1 +
x1))), None, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (15 + -1 * tl_math.abs(-15 + tl_math.abs(-1 +
x0))), None, eviction_policy='evict_last')
tmp10 = tl.load(in_ptr2 + x2, None, eviction_policy='evict_last')
tmp1 = tl.full([XBLOCK], 8, tl.int32)
tmp2 = tmp0 + tmp1
tmp3 = tmp0 < 0
tmp4 = tl.where(tmp3, tmp2, tmp0)
tmp6 = tmp5 + tmp1
tmp7 = tmp5 < 0
tmp8 = tl.where(tmp7, tmp6, tmp5)
tmp9 = tl.load(in_ptr1 + (tmp8 + 8 * tmp4 + 64 * x4), None,
eviction_policy='evict_last')
tmp11 = tmp9 + tmp10
tmp12 = tl.full([1], 0, tl.int32)
tmp13 = triton_helpers.maximum(tmp12, tmp11)
tl.store(out_ptr0 + x7, tmp13, None)
@triton.jit
def triton_poi_fused_convolution_reflection_pad2d_relu_6(in_ptr0, in_ptr1,
out_ptr0, xnumel, XBLOCK: tl.constexpr):
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
tl.full([XBLOCK], True, tl.int1)
x0 = xindex % 18
x1 = xindex // 18 % 18
x4 = xindex // 324
x2 = xindex // 324 % 256
x5 = xindex
tmp0 = tl.load(in_ptr0 + (255 + -1 * tl_math.abs(-15 + tl_math.abs(-1 +
x0)) + -16 * tl_math.abs(-15 + tl_math.abs(-1 + x1)) + 256 * x4),
None, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + x2, None, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tl.store(out_ptr0 + x5, tmp4, None)
@triton.jit
def triton_poi_fused__to_copy_add_arange_mul_7(out_ptr0, xnumel, XBLOCK: tl
.constexpr):
xnumel = 32
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = x0
tmp1 = tmp0.to(tl.float32)
tmp2 = 0.5
tmp3 = tmp1 * tmp2
tmp4 = tmp3.to(tl.int32)
tl.store(out_ptr0 + x0, tmp4, xmask)
@triton.jit
def triton_poi_fused__unsafe_index_convolution_reflection_pad2d_relu_8(in_ptr0,
in_ptr1, in_ptr2, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
tl.full([XBLOCK], True, tl.int1)
x1 = xindex // 34 % 34
x0 = xindex % 34
x4 = xindex // 1156
x2 = xindex // 1156 % 128
x7 = xindex
tmp0 = tl.load(in_ptr0 + (31 + -1 * tl_math.abs(-31 + tl_math.abs(-1 +
x1))), None, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (31 + -1 * tl_math.abs(-31 + tl_math.abs(-1 +
x0))), None, eviction_policy='evict_last')
tmp10 = tl.load(in_ptr2 + x2, None, eviction_policy='evict_last')
tmp1 = tl.full([XBLOCK], 16, tl.int32)
tmp2 = tmp0 + tmp1
tmp3 = tmp0 < 0
tmp4 = tl.where(tmp3, tmp2, tmp0)
tmp6 = tmp5 + tmp1
tmp7 = tmp5 < 0
tmp8 = tl.where(tmp7, tmp6, tmp5)
tmp9 = tl.load(in_ptr1 + (tmp8 + 16 * tmp4 + 256 * x4), None,
eviction_policy='evict_last')
tmp11 = tmp9 + tmp10
tmp12 = tl.full([1], 0, tl.int32)
tmp13 = triton_helpers.maximum(tmp12, tmp11)
tl.store(out_ptr0 + x7, tmp13, None)
@triton.jit
def triton_poi_fused_convolution_reflection_pad2d_relu_9(in_ptr0, in_ptr1,
out_ptr0, xnumel, XBLOCK: tl.constexpr):
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
tl.full([XBLOCK], True, tl.int1)
x0 = xindex % 34
x1 = xindex // 34 % 34
x4 = xindex // 1156
x2 = xindex // 1156 % 128
x5 = xindex
tmp0 = tl.load(in_ptr0 + (1023 + -1 * tl_math.abs(-31 + tl_math.abs(-1 +
x0)) + -32 * tl_math.abs(-31 + tl_math.abs(-1 + x1)) + 1024 * x4),
None, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + x2, None, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tl.store(out_ptr0 + x5, tmp4, None)
@triton.jit
def triton_poi_fused_arange_10(out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = x0
tl.store(out_ptr0 + x0, tmp0, xmask)
@triton.jit
def triton_poi_fused__to_copy_add_arange_mul_11(out_ptr0, xnumel, XBLOCK:
tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = x0
tmp1 = tmp0.to(tl.float32)
tmp2 = 0.5
tmp3 = tmp1 * tmp2
tmp4 = tmp3.to(tl.int32)
tl.store(out_ptr0 + x0, tmp4, xmask)
@triton.jit
def triton_poi_fused__unsafe_index_convolution_reflection_pad2d_relu_12(in_ptr0
, in_ptr1, in_ptr2, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 1115136
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = xindex // 66 % 66
x0 = xindex % 66
x4 = xindex // 4356
x2 = xindex // 4356 % 64
x7 = xindex
tmp0 = tl.load(in_ptr0 + (63 + -1 * tl_math.abs(-63 + tl_math.abs(-1 +
x1))), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (63 + -1 * tl_math.abs(-63 + tl_math.abs(-1 +
x0))), xmask, eviction_policy='evict_last')
tmp10 = tl.load(in_ptr2 + x2, xmask, eviction_policy='evict_last')
tmp1 = tl.full([XBLOCK], 32, tl.int32)
tmp2 = tmp0 + tmp1
tmp3 = tmp0 < 0
tmp4 = tl.where(tmp3, tmp2, tmp0)
tmp6 = tmp5 + tmp1
tmp7 = tmp5 < 0
tmp8 = tl.where(tmp7, tmp6, tmp5)
tmp9 = tl.load(in_ptr1 + (tmp8 + 32 * tmp4 + 1024 * x4), xmask,
eviction_policy='evict_last')
tmp11 = tmp9 + tmp10
tmp12 = tl.full([1], 0, tl.int32)
tmp13 = triton_helpers.maximum(tmp12, tmp11)
tl.store(out_ptr0 + x7, tmp13, xmask)
@triton.jit
def triton_poi_fused_convolution_reflection_pad2d_relu_13(in_ptr0, in_ptr1,
out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 1115136
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 66
x1 = xindex // 66 % 66
x4 = xindex // 4356
x2 = xindex // 4356 % 64
x5 = xindex
tmp0 = tl.load(in_ptr0 + (4095 + -1 * tl_math.abs(-63 + tl_math.abs(-1 +
x0)) + -64 * tl_math.abs(-63 + tl_math.abs(-1 + x1)) + 4096 * x4),
xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + x2, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tl.store(out_ptr0 + x5, tmp4, xmask)
@triton.jit
def triton_poi_fused_convolution_14(in_out_ptr0, in_ptr0, xnumel, XBLOCK:
tl.constexpr):
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
tl.full([XBLOCK], True, tl.int1)
x3 = xindex
x1 = xindex // 4096 % 3
tmp0 = tl.load(in_out_ptr0 + x3, None)
tmp1 = tl.load(in_ptr0 + x1, None, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + x3, tmp2, None)
@triton.jit
def triton_poi_fused_convolution_relu_threshold_backward_15(in_ptr0,
in_ptr1, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
tl.full([XBLOCK], True, tl.int1)
x3 = xindex
x1 = xindex // 4096 % 64
tmp0 = tl.load(in_ptr0 + x3, None)
tmp1 = tl.load(in_ptr1 + x1, None, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(out_ptr0 + x3, tmp6, None)
@triton.jit
def triton_poi_fused_convolution_relu_threshold_backward_16(in_ptr0,
in_ptr1, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
tl.full([XBLOCK], True, tl.int1)
x3 = xindex
x1 = xindex // 1024 % 64
tmp0 = tl.load(in_ptr0 + x3, None)
tmp1 = tl.load(in_ptr1 + x1, None, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(out_ptr0 + x3, tmp6, None)
@triton.jit
def triton_poi_fused_convolution_relu_threshold_backward_17(in_ptr0,
in_ptr1, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
tl.full([XBLOCK], True, tl.int1)
x3 = xindex
x1 = xindex // 1024 % 128
tmp0 = tl.load(in_ptr0 + x3, None)
tmp1 = tl.load(in_ptr1 + x1, None, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(out_ptr0 + x3, tmp6, None)
@triton.jit
def triton_poi_fused_convolution_relu_threshold_backward_18(in_ptr0,
in_ptr1, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
tl.full([XBLOCK], True, tl.int1)
x3 = xindex
x1 = xindex // 256 % 128
tmp0 = tl.load(in_ptr0 + x3, None)
tmp1 = tl.load(in_ptr1 + x1, None, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(out_ptr0 + x3, tmp6, None)
@triton.jit
def triton_poi_fused_convolution_relu_threshold_backward_19(in_ptr0,
in_ptr1, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
tl.full([XBLOCK], True, tl.int1)
x3 = xindex
x1 = xindex // 256 % 256
tmp0 = tl.load(in_ptr0 + x3, None)
tmp1 = tl.load(in_ptr1 + x1, None, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(out_ptr0 + x3, tmp6, None)
@triton.jit
def triton_poi_fused_convolution_relu_threshold_backward_20(in_ptr0,
in_ptr1, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
tl.full([XBLOCK], True, tl.int1)
x3 = xindex
x1 = xindex // 64 % 256
tmp0 = tl.load(in_ptr0 + x3, None)
tmp1 = tl.load(in_ptr1 + x1, None, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(out_ptr0 + x3, tmp6, None)
@triton.jit
def triton_poi_fused_convolution_relu_threshold_backward_21(in_ptr0,
in_ptr1, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
tl.full([XBLOCK], True, tl.int1)
x3 = xindex
x1 = xindex // 64 % 512
tmp0 = tl.load(in_ptr0 + x3, None)
tmp1 = tl.load(in_ptr1 + x1, None, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(out_ptr0 + x3, tmp6, None)
@triton.jit
def triton_poi_fused_convolution_relu_threshold_backward_22(in_ptr0,
in_ptr1, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
tl.full([XBLOCK], True, tl.int1)
x3 = xindex
x1 = xindex // 16 % 512
tmp0 = tl.load(in_ptr0 + x3, None)
tmp1 = tl.load(in_ptr1 + x1, None, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(out_ptr0 + x3, tmp6, None)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7, primals_8, primals_9, primals_10, primals_11, primals_12,
primals_13, primals_14, primals_15, primals_16, primals_17,
primals_18, primals_19, primals_20, primals_21, primals_22,
primals_23, primals_24, primals_25, primals_26, primals_27) = args
args.clear()
assert_size_stride(primals_1, (4, 512, 4, 4), (8192, 16, 4, 1))
assert_size_stride(primals_2, (512, 512, 3, 3), (4608, 9, 3, 1))
assert_size_stride(primals_3, (512,), (1,))
assert_size_stride(primals_4, (512, 512, 3, 3), (4608, 9, 3, 1))
assert_size_stride(primals_5, (512,), (1,))
assert_size_stride(primals_6, (512, 512, 3, 3), (4608, 9, 3, 1))
assert_size_stride(primals_7, (512,), (1,))
assert_size_stride(primals_8, (512, 512, 3, 3), (4608, 9, 3, 1))
assert_size_stride(primals_9, (512,), (1,))
assert_size_stride(primals_10, (256, 512, 3, 3), (4608, 9, 3, 1))
assert_size_stride(primals_11, (256,), (1,))
assert_size_stride(primals_12, (256, 256, 3, 3), (2304, 9, 3, 1))
assert_size_stride(primals_13, (256,), (1,))
assert_size_stride(primals_14, (256, 256, 3, 3), (2304, 9, 3, 1))
assert_size_stride(primals_15, (256,), (1,))
assert_size_stride(primals_16, (256, 256, 3, 3), (2304, 9, 3, 1))
assert_size_stride(primals_17, (256,), (1,))
assert_size_stride(primals_18, (128, 256, 3, 3), (2304, 9, 3, 1))
assert_size_stride(primals_19, (128,), (1,))
assert_size_stride(primals_20, (128, 128, 3, 3), (1152, 9, 3, 1))
assert_size_stride(primals_21, (128,), (1,))
assert_size_stride(primals_22, (64, 128, 3, 3), (1152, 9, 3, 1))
assert_size_stride(primals_23, (64,), (1,))
assert_size_stride(primals_24, (64, 64, 3, 3), (576, 9, 3, 1))
assert_size_stride(primals_25, (64,), (1,))
assert_size_stride(primals_26, (3, 64, 3, 3), (576, 9, 3, 1))
assert_size_stride(primals_27, (3,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 512, 6, 6), (18432, 36, 6, 1), torch.
float32)
get_raw_stream(0)
triton_poi_fused_reflection_pad2d_0[grid(73728)](primals_1, buf0,
73728, XBLOCK=1024, num_warps=4, num_stages=1)
del primals_1
buf1 = extern_kernels.convolution(buf0, primals_2, stride=(1, 1),
padding=(0, 0), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf1, (4, 512, 4, 4), (8192, 16, 4, 1))
buf2 = empty_strided_cuda((8,), (1,), torch.int64)
triton_poi_fused__to_copy_add_arange_mul_1[grid(8)](buf2, 8, XBLOCK
=8, num_warps=1, num_stages=1)
buf3 = empty_strided_cuda((4, 512, 10, 10), (51200, 100, 10, 1),
torch.float32)
triton_poi_fused__unsafe_index_convolution_reflection_pad2d_relu_2[grid
(204800)](buf2, buf1, primals_3, buf3, 204800, XBLOCK=512,
num_warps=8, num_stages=1)
buf4 = extern_kernels.convolution(buf3, primals_4, stride=(1, 1),
padding=(0, 0), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf4, (4, 512, 8, 8), (32768, 64, 8, 1))
buf5 = empty_strided_cuda((4, 512, 10, 10), (51200, 100, 10, 1),
torch.float32)
triton_poi_fused_convolution_reflection_pad2d_relu_3[grid(204800)](buf4
, primals_5, buf5, 204800, XBLOCK=512, num_warps=8, num_stages=1)
buf6 = extern_kernels.convolution(buf5, primals_6, stride=(1, 1),
padding=(0, 0), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf6, (4, 512, 8, 8), (32768, 64, 8, 1))
buf7 = empty_strided_cuda((4, 512, 10, 10), (51200, 100, 10, 1),
torch.float32)
triton_poi_fused_convolution_reflection_pad2d_relu_3[grid(204800)](buf6
, primals_7, buf7, 204800, XBLOCK=512, num_warps=8, num_stages=1)
buf8 = extern_kernels.convolution(buf7, primals_8, stride=(1, 1),
padding=(0, 0), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf8, (4, 512, 8, 8), (32768, 64, 8, 1))
buf9 = empty_strided_cuda((4, 512, 10, 10), (51200, 100, 10, 1),
torch.float32)
triton_poi_fused_convolution_reflection_pad2d_relu_3[grid(204800)](buf8
, primals_9, buf9, 204800, XBLOCK=512, num_warps=8, num_stages=1)
buf10 = extern_kernels.convolution(buf9, primals_10, stride=(1, 1),
padding=(0, 0), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf10, (4, 256, 8, 8), (16384, 64, 8, 1))
buf11 = empty_strided_cuda((16,), (1,), torch.int64)
triton_poi_fused__to_copy_add_arange_mul_4[grid(16)](buf11, 16,
XBLOCK=16, num_warps=1, num_stages=1)
buf12 = empty_strided_cuda((4, 256, 18, 18), (82944, 324, 18, 1),
torch.float32)
triton_poi_fused__unsafe_index_convolution_reflection_pad2d_relu_5[grid
(331776)](buf11, buf10, primals_11, buf12, 331776, XBLOCK=512,
num_warps=8, num_stages=1)
buf13 = extern_kernels.convolution(buf12, primals_12, stride=(1, 1),
padding=(0, 0), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf13, (4, 256, 16, 16), (65536, 256, 16, 1))
buf14 = empty_strided_cuda((4, 256, 18, 18), (82944, 324, 18, 1),
torch.float32)
triton_poi_fused_convolution_reflection_pad2d_relu_6[grid(331776)](
buf13, primals_13, buf14, 331776, XBLOCK=512, num_warps=8,
num_stages=1)
buf15 = extern_kernels.convolution(buf14, primals_14, stride=(1, 1),
padding=(0, 0), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf15, (4, 256, 16, 16), (65536, 256, 16, 1))
buf16 = empty_strided_cuda((4, 256, 18, 18), (82944, 324, 18, 1),
torch.float32)
triton_poi_fused_convolution_reflection_pad2d_relu_6[grid(331776)](
buf15, primals_15, buf16, 331776, XBLOCK=512, num_warps=8,
num_stages=1)
buf17 = extern_kernels.convolution(buf16, primals_16, stride=(1, 1),
padding=(0, 0), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf17, (4, 256, 16, 16), (65536, 256, 16, 1))
buf18 = empty_strided_cuda((4, 256, 18, 18), (82944, 324, 18, 1),
torch.float32)
triton_poi_fused_convolution_reflection_pad2d_relu_6[grid(331776)](
buf17, primals_17, buf18, 331776, XBLOCK=512, num_warps=8,
num_stages=1)
buf19 = extern_kernels.convolution(buf18, primals_18, stride=(1, 1),
padding=(0, 0), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf19, (4, 128, 16, 16), (32768, 256, 16, 1))
buf20 = empty_strided_cuda((32,), (1,), torch.int64)
triton_poi_fused__to_copy_add_arange_mul_7[grid(32)](buf20, 32,
XBLOCK=32, num_warps=1, num_stages=1)
buf21 = empty_strided_cuda((4, 128, 34, 34), (147968, 1156, 34, 1),
torch.float32)
triton_poi_fused__unsafe_index_convolution_reflection_pad2d_relu_8[grid
(591872)](buf20, buf19, primals_19, buf21, 591872, XBLOCK=1024,
num_warps=4, num_stages=1)
buf22 = extern_kernels.convolution(buf21, primals_20, stride=(1, 1),
padding=(0, 0), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf22, (4, 128, 32, 32), (131072, 1024, 32, 1))
buf23 = empty_strided_cuda((4, 128, 34, 34), (147968, 1156, 34, 1),
torch.float32)
triton_poi_fused_convolution_reflection_pad2d_relu_9[grid(591872)](
buf22, primals_21, buf23, 591872, XBLOCK=512, num_warps=8,
num_stages=1)
buf24 = extern_kernels.convolution(buf23, primals_22, stride=(1, 1),
padding=(0, 0), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf24, (4, 64, 32, 32), (65536, 1024, 32, 1))
buf25 = empty_strided_cuda((64,), (1,), torch.int64)
triton_poi_fused_arange_10[grid(64)](buf25, 64, XBLOCK=64,
num_warps=1, num_stages=1)
buf26 = empty_strided_cuda((64,), (1,), torch.int64)
triton_poi_fused__to_copy_add_arange_mul_11[grid(64)](buf26, 64,
XBLOCK=64, num_warps=1, num_stages=1)
buf27 = empty_strided_cuda((4, 64, 66, 66), (278784, 4356, 66, 1),
torch.float32)
triton_poi_fused__unsafe_index_convolution_reflection_pad2d_relu_12[
grid(1115136)](buf26, buf24, primals_23, buf27, 1115136, XBLOCK
=512, num_warps=8, num_stages=1)
buf28 = extern_kernels.convolution(buf27, primals_24, stride=(1, 1),
padding=(0, 0), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf28, (4, 64, 64, 64), (262144, 4096, 64, 1))
buf29 = empty_strided_cuda((4, 64, 66, 66), (278784, 4356, 66, 1),
torch.float32)
triton_poi_fused_convolution_reflection_pad2d_relu_13[grid(1115136)](
buf28, primals_25, buf29, 1115136, XBLOCK=512, num_warps=8,
num_stages=1)
buf30 = extern_kernels.convolution(buf29, primals_26, stride=(1, 1),
padding=(0, 0), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf30, (4, 3, 64, 64), (12288, 4096, 64, 1))
buf31 = buf30
del buf30
triton_poi_fused_convolution_14[grid(49152)](buf31, primals_27,
49152, XBLOCK=256, num_warps=4, num_stages=1)
del primals_27
buf32 = empty_strided_cuda((4, 64, 64, 64), (262144, 4096, 64, 1),
torch.bool)
triton_poi_fused_convolution_relu_threshold_backward_15[grid(1048576)](
buf28, primals_25, buf32, 1048576, XBLOCK=1024, num_warps=4,
num_stages=1)
del buf28
del primals_25
buf33 = empty_strided_cuda((4, 64, 32, 32), (65536, 1024, 32, 1),
torch.bool)
triton_poi_fused_convolution_relu_threshold_backward_16[grid(262144)](
buf24, primals_23, buf33, 262144, XBLOCK=512, num_warps=8,
num_stages=1)
del buf24
del primals_23
buf34 = empty_strided_cuda((4, 128, 32, 32), (131072, 1024, 32, 1),
torch.bool)
triton_poi_fused_convolution_relu_threshold_backward_17[grid(524288)](
buf22, primals_21, buf34, 524288, XBLOCK=1024, num_warps=4,
num_stages=1)
del buf22
del primals_21
buf35 = empty_strided_cuda((4, 128, 16, 16), (32768, 256, 16, 1),
torch.bool)
triton_poi_fused_convolution_relu_threshold_backward_18[grid(131072)](
buf19, primals_19, buf35, 131072, XBLOCK=1024, num_warps=4,
num_stages=1)
del buf19
del primals_19
buf36 = empty_strided_cuda((4, 256, 16, 16), (65536, 256, 16, 1),
torch.bool)
triton_poi_fused_convolution_relu_threshold_backward_19[grid(262144)](
buf17, primals_17, buf36, 262144, XBLOCK=1024, num_warps=4,
num_stages=1)
del buf17
del primals_17
buf37 = empty_strided_cuda((4, 256, 16, 16), (65536, 256, 16, 1),
torch.bool)
triton_poi_fused_convolution_relu_threshold_backward_19[grid(262144)](
buf15, primals_15, buf37, 262144, XBLOCK=1024, num_warps=4,
num_stages=1)
del buf15
del primals_15
buf38 = empty_strided_cuda((4, 256, 16, 16), (65536, 256, 16, 1),
torch.bool)
triton_poi_fused_convolution_relu_threshold_backward_19[grid(262144)](
buf13, primals_13, buf38, 262144, XBLOCK=1024, num_warps=4,
num_stages=1)
del buf13
del primals_13
buf39 = empty_strided_cuda((4, 256, 8, 8), (16384, 64, 8, 1), torch
.bool)
triton_poi_fused_convolution_relu_threshold_backward_20[grid(65536)](
buf10, primals_11, buf39, 65536, XBLOCK=256, num_warps=4,
num_stages=1)
del buf10
del primals_11
buf40 = empty_strided_cuda((4, 512, 8, 8), (32768, 64, 8, 1), torch
.bool)
triton_poi_fused_convolution_relu_threshold_backward_21[grid(131072)](
buf8, primals_9, buf40, 131072, XBLOCK=1024, num_warps=4,
num_stages=1)
del buf8
del primals_9
buf41 = empty_strided_cuda((4, 512, 8, 8), (32768, 64, 8, 1), torch
.bool)
triton_poi_fused_convolution_relu_threshold_backward_21[grid(131072)](
buf6, primals_7, buf41, 131072, XBLOCK=1024, num_warps=4,
num_stages=1)
del buf6
del primals_7
buf42 = empty_strided_cuda((4, 512, 8, 8), (32768, 64, 8, 1), torch
.bool)
triton_poi_fused_convolution_relu_threshold_backward_21[grid(131072)](
buf4, primals_5, buf42, 131072, XBLOCK=1024, num_warps=4,
num_stages=1)
del buf4
del primals_5
buf43 = empty_strided_cuda((4, 512, 4, 4), (8192, 16, 4, 1), torch.bool
)
triton_poi_fused_convolution_relu_threshold_backward_22[grid(32768)](
buf1, primals_3, buf43, 32768, XBLOCK=256, num_warps=4,
num_stages=1)
del buf1
del primals_3
return (buf31, primals_2, primals_4, primals_6, primals_8, primals_10,
primals_12, primals_14, primals_16, primals_18, primals_20,
primals_22, primals_24, primals_26, buf0, buf2, buf3, buf5, buf7,
buf9, buf11, buf12, buf14, buf16, buf18, buf20, buf21, buf23, buf25,
buf26, buf27, buf29, buf32, buf33, buf34, buf35, buf36, buf37,
buf38, buf39, buf40, buf41, buf42, buf43)
class Decoder5New(nn.Module):
def __init__(self):
super(Decoder5New, self).__init__()
self.reflecPad15 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv15 = nn.Conv2d(512, 512, 3, 1, 0)
self.relu15 = nn.ReLU(inplace=True)
self.unpool = nn.UpsamplingNearest2d(scale_factor=2)
self.reflecPad16 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv16 = nn.Conv2d(512, 512, 3, 1, 0)
self.relu16 = nn.ReLU(inplace=True)
self.reflecPad17 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv17 = nn.Conv2d(512, 512, 3, 1, 0)
self.relu17 = nn.ReLU(inplace=True)
self.reflecPad18 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv18 = nn.Conv2d(512, 512, 3, 1, 0)
self.relu18 = nn.ReLU(inplace=True)
self.reflecPad19 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv19 = nn.Conv2d(512, 256, 3, 1, 0)
self.relu19 = nn.ReLU(inplace=True)
self.unpool2 = nn.UpsamplingNearest2d(scale_factor=2)
self.reflecPad20 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv20 = nn.Conv2d(256, 256, 3, 1, 0)
self.relu20 = nn.ReLU(inplace=True)
self.reflecPad21 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv21 = nn.Conv2d(256, 256, 3, 1, 0)
self.relu21 = nn.ReLU(inplace=True)
self.reflecPad22 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv22 = nn.Conv2d(256, 256, 3, 1, 0)
self.relu22 = nn.ReLU(inplace=True)
self.reflecPad23 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv23 = nn.Conv2d(256, 128, 3, 1, 0)
self.relu23 = nn.ReLU(inplace=True)
self.unpool3 = nn.UpsamplingNearest2d(scale_factor=2)
self.reflecPad24 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv24 = nn.Conv2d(128, 128, 3, 1, 0)
self.relu24 = nn.ReLU(inplace=True)
self.reflecPad25 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv25 = nn.Conv2d(128, 64, 3, 1, 0)
self.relu25 = nn.ReLU(inplace=True)
self.unpool4 = nn.UpsamplingNearest2d(scale_factor=2)
self.reflecPad26 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv26 = nn.Conv2d(64, 64, 3, 1, 0)
self.relu26 = nn.ReLU(inplace=True)
self.reflecPad27 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv27 = nn.Conv2d(64, 3, 3, 1, 0)
def forward(self, input_0):
primals_2 = self.conv15.weight
primals_3 = self.conv15.bias
primals_4 = self.conv16.weight
primals_5 = self.conv16.bias
primals_6 = self.conv17.weight
primals_7 = self.conv17.bias
primals_8 = self.conv18.weight
primals_9 = self.conv18.bias
primals_10 = self.conv19.weight
primals_11 = self.conv19.bias
primals_12 = self.conv20.weight
primals_13 = self.conv20.bias
primals_14 = self.conv21.weight
primals_15 = self.conv21.bias
primals_16 = self.conv22.weight
primals_17 = self.conv22.bias
primals_18 = self.conv23.weight
primals_19 = self.conv23.bias
primals_20 = self.conv24.weight
primals_21 = self.conv24.bias
primals_22 = self.conv25.weight
primals_23 = self.conv25.bias
primals_24 = self.conv26.weight
primals_25 = self.conv26.bias
primals_26 = self.conv27.weight
primals_27 = self.conv27.bias
primals_1 = input_0
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7, primals_8, primals_9,
primals_10, primals_11, primals_12, primals_13, primals_14,
primals_15, primals_16, primals_17, primals_18, primals_19,
primals_20, primals_21, primals_22, primals_23, primals_24,
primals_25, primals_26, primals_27])
return output[0]
| hologerry/wct_experiment | Decoder5 | false | 6,897 | [
"MIT"
] | 1 | 890d885561dc8df8c4ae732aebd902aa838257e6 | https://github.com/hologerry/wct_experiment/tree/890d885561dc8df8c4ae732aebd902aa838257e6 | import torch
import torch.nn as nn
class Model(nn.Module):
def __init__(self):
super().__init__()
self.reflecPad15 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv15 = nn.Conv2d(512, 512, 3, 1, 0)
self.relu15 = nn.ReLU(inplace=True)
self.unpool = nn.UpsamplingNearest2d(scale_factor=2)
self.reflecPad16 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv16 = nn.Conv2d(512, 512, 3, 1, 0)
self.relu16 = nn.ReLU(inplace=True)
self.reflecPad17 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv17 = nn.Conv2d(512, 512, 3, 1, 0)
self.relu17 = nn.ReLU(inplace=True)
self.reflecPad18 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv18 = nn.Conv2d(512, 512, 3, 1, 0)
self.relu18 = nn.ReLU(inplace=True)
self.reflecPad19 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv19 = nn.Conv2d(512, 256, 3, 1, 0)
self.relu19 = nn.ReLU(inplace=True)
self.unpool2 = nn.UpsamplingNearest2d(scale_factor=2)
self.reflecPad20 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv20 = nn.Conv2d(256, 256, 3, 1, 0)
self.relu20 = nn.ReLU(inplace=True)
self.reflecPad21 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv21 = nn.Conv2d(256, 256, 3, 1, 0)
self.relu21 = nn.ReLU(inplace=True)
self.reflecPad22 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv22 = nn.Conv2d(256, 256, 3, 1, 0)
self.relu22 = nn.ReLU(inplace=True)
self.reflecPad23 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv23 = nn.Conv2d(256, 128, 3, 1, 0)
self.relu23 = nn.ReLU(inplace=True)
self.unpool3 = nn.UpsamplingNearest2d(scale_factor=2)
self.reflecPad24 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv24 = nn.Conv2d(128, 128, 3, 1, 0)
self.relu24 = nn.ReLU(inplace=True)
self.reflecPad25 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv25 = nn.Conv2d(128, 64, 3, 1, 0)
self.relu25 = nn.ReLU(inplace=True)
self.unpool4 = nn.UpsamplingNearest2d(scale_factor=2)
self.reflecPad26 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv26 = nn.Conv2d(64, 64, 3, 1, 0)
self.relu26 = nn.ReLU(inplace=True)
self.reflecPad27 = nn.ReflectionPad2d((1, 1, 1, 1))
self.conv27 = nn.Conv2d(64, 3, 3, 1, 0)
def forward(self, x):
out = self.reflecPad15(x)
out = self.conv15(out)
out = self.relu15(out)
out = self.unpool(out)
out = self.reflecPad16(out)
out = self.conv16(out)
out = self.relu16(out)
out = self.reflecPad17(out)
out = self.conv17(out)
out = self.relu17(out)
out = self.reflecPad18(out)
out = self.conv18(out)
out = self.relu18(out)
out = self.reflecPad19(out)
out = self.conv19(out)
out = self.relu19(out)
out = self.unpool2(out)
out = self.reflecPad20(out)
out = self.conv20(out)
out = self.relu20(out)
out = self.reflecPad21(out)
out = self.conv21(out)
out = self.relu21(out)
out = self.reflecPad22(out)
out = self.conv22(out)
out = self.relu22(out)
out = self.reflecPad23(out)
out = self.conv23(out)
out = self.relu23(out)
out = self.unpool3(out)
out = self.reflecPad24(out)
out = self.conv24(out)
out = self.relu24(out)
out = self.reflecPad25(out)
out = self.conv25(out)
out = self.relu25(out)
out = self.unpool4(out)
out = self.reflecPad26(out)
out = self.conv26(out)
out = self.relu26(out)
out = self.reflecPad27(out)
out = self.conv27(out)
return out
def get_inputs():
return [torch.rand([4, 512, 4, 4])]
def get_init_inputs():
return []
|
FirstLSTMAmp | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/pr/cprthrqz6iotcmrjfcrj7taqntzxisdcjtr54gsuz2ck2kf6kbsr.py
# Topologically Sorted Source Nodes: [x_1], Original ATen: [aten.relu, aten.threshold_backward]
# Source node to ATen node mapping:
# x_1 => relu
# Graph fragment:
# %relu : [num_users=1] = call_function[target=torch.ops.aten.relu.default](args = (%view_1,), kwargs = {})
# %le : [num_users=1] = call_function[target=torch.ops.aten.le.Scalar](args = (%view_6, 0), kwargs = {})
triton_poi_fused_relu_threshold_backward_0 = async_compile.triton('triton_poi_fused_relu_threshold_backward_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*i1', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_relu_threshold_backward_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_relu_threshold_backward_0(in_out_ptr0, in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_out_ptr0 + (x0), xmask)
tmp1 = tl.load(in_ptr0 + (0))
tmp2 = tl.broadcast_to(tmp1, [XBLOCK])
tmp3 = tmp0 + tmp2
tmp4 = tl.full([1], 0, tl.int32)
tmp5 = triton_helpers.maximum(tmp4, tmp3)
tmp6 = 0.0
tmp7 = tmp5 <= tmp6
tl.store(in_out_ptr0 + (x0), tmp5, xmask)
tl.store(out_ptr0 + (x0), tmp7, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/3f/c3fojdylt3acgtptumtowi32d26uvrqtyzefg7ta476ezqj5dtyp.py
# Topologically Sorted Source Nodes: [x_2], Original ATen: [aten.view]
# Source node to ATen node mapping:
# x_2 => view_7
# Graph fragment:
# %view_7 : [num_users=2] = call_function[target=torch.ops.aten.reshape.default](args = (%view_6, [64, 1]), kwargs = {})
triton_poi_fused_view_1 = async_compile.triton('triton_poi_fused_view_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_view_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_view_1(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (x0 + (4*((x0 % 4) // 4)) + (16*(((4*((x0 // 4) % 4)) + (x0 % 4)) // 16))), xmask)
tl.store(out_ptr0 + (x0), tmp0, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5 = args
args.clear()
assert_size_stride(primals_1, (1, 4), (4, 1))
assert_size_stride(primals_2, (1, ), (1, ))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_4, (4, 1), (1, 1))
assert_size_stride(primals_5, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 1), (1, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(primals_3, (64, 4), (4, 1), 0), reinterpret_tensor(primals_1, (4, 1), (1, 4), 0), out=buf0)
del primals_1
buf1 = reinterpret_tensor(buf0, (4, 4, 4, 1), (16, 4, 1, 1), 0); del buf0 # reuse
buf4 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 1), torch.bool)
# Topologically Sorted Source Nodes: [x_1], Original ATen: [aten.relu, aten.threshold_backward]
stream0 = get_raw_stream(0)
triton_poi_fused_relu_threshold_backward_0.run(buf1, primals_2, buf4, 64, grid=grid(64), stream=stream0)
del primals_2
buf2 = empty_strided_cuda((64, 1), (1, 1), torch.float32)
# Topologically Sorted Source Nodes: [x_2], Original ATen: [aten.view]
triton_poi_fused_view_1.run(buf1, buf2, 64, grid=grid(64), stream=stream0)
del buf1
buf3 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [x_2], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_5, buf2, reinterpret_tensor(primals_4, (1, 4), (1, 1), 0), alpha=1, beta=1, out=buf3)
del primals_5
return (reinterpret_tensor(buf3, (4, 4, 4, 4), (64, 16, 4, 1), 0), reinterpret_tensor(primals_3, (64, 4), (4, 1), 0), buf2, primals_4, buf4, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((1, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((1, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, 1), (1, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
class FirstLSTMAmp(nn.Module):
"""
First LSTM amplifier branch.
Parameters:
----------
in_features : int
Number of input channels.
out_features : int
Number of output channels.
"""
def __init__(self, in_features, out_features):
super(FirstLSTMAmp, self).__init__()
mid_features = in_features // 4
self.fc1 = nn.Linear(in_features=in_features, out_features=mid_features
)
self.activ = nn.ReLU(inplace=True)
self.fc2 = nn.Linear(in_features=mid_features, out_features=
out_features)
def forward(self, x):
x = self.fc1(x)
x = self.activ(x)
x = self.fc2(x)
return x
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'in_features': 4, 'out_features': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_relu_threshold_backward_0(in_out_ptr0, in_ptr0,
out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_out_ptr0 + x0, xmask)
tmp1 = tl.load(in_ptr0 + 0)
tmp2 = tl.broadcast_to(tmp1, [XBLOCK])
tmp3 = tmp0 + tmp2
tmp4 = tl.full([1], 0, tl.int32)
tmp5 = triton_helpers.maximum(tmp4, tmp3)
tmp6 = 0.0
tmp7 = tmp5 <= tmp6
tl.store(in_out_ptr0 + x0, tmp5, xmask)
tl.store(out_ptr0 + x0, tmp7, xmask)
@triton.jit
def triton_poi_fused_view_1(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (x0 + 4 * (x0 % 4 // 4) + 16 * ((4 * (x0 // 4 %
4) + x0 % 4) // 16)), xmask)
tl.store(out_ptr0 + x0, tmp0, xmask)
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5 = args
args.clear()
assert_size_stride(primals_1, (1, 4), (4, 1))
assert_size_stride(primals_2, (1,), (1,))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_4, (4, 1), (1, 1))
assert_size_stride(primals_5, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 1), (1, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_3, (64, 4), (4, 1), 0),
reinterpret_tensor(primals_1, (4, 1), (1, 4), 0), out=buf0)
del primals_1
buf1 = reinterpret_tensor(buf0, (4, 4, 4, 1), (16, 4, 1, 1), 0)
del buf0
buf4 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 1), torch.bool)
get_raw_stream(0)
triton_poi_fused_relu_threshold_backward_0[grid(64)](buf1,
primals_2, buf4, 64, XBLOCK=64, num_warps=1, num_stages=1)
del primals_2
buf2 = empty_strided_cuda((64, 1), (1, 1), torch.float32)
triton_poi_fused_view_1[grid(64)](buf1, buf2, 64, XBLOCK=64,
num_warps=1, num_stages=1)
del buf1
buf3 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
extern_kernels.addmm(primals_5, buf2, reinterpret_tensor(primals_4,
(1, 4), (1, 1), 0), alpha=1, beta=1, out=buf3)
del primals_5
return reinterpret_tensor(buf3, (4, 4, 4, 4), (64, 16, 4, 1), 0
), reinterpret_tensor(primals_3, (64, 4), (4, 1), 0
), buf2, primals_4, buf4
class FirstLSTMAmpNew(nn.Module):
"""
First LSTM amplifier branch.
Parameters:
----------
in_features : int
Number of input channels.
out_features : int
Number of output channels.
"""
def __init__(self, in_features, out_features):
super(FirstLSTMAmpNew, self).__init__()
mid_features = in_features // 4
self.fc1 = nn.Linear(in_features=in_features, out_features=mid_features
)
self.activ = nn.ReLU(inplace=True)
self.fc2 = nn.Linear(in_features=mid_features, out_features=
out_features)
def forward(self, input_0):
primals_1 = self.fc1.weight
primals_2 = self.fc1.bias
primals_4 = self.fc2.weight
primals_5 = self.fc2.bias
primals_3 = input_0
output = call([primals_1, primals_2, primals_3, primals_4, primals_5])
return output[0]
| iofthetiger/pkuad | FirstLSTMAmp | false | 6,898 | [
"Apache-2.0"
] | 1 | 07496d108c614c84be028f344830becc9cac8fe5 | https://github.com/iofthetiger/pkuad/tree/07496d108c614c84be028f344830becc9cac8fe5 | import torch
import torch.nn as nn
class Model(nn.Module):
"""
First LSTM amplifier branch.
Parameters:
----------
in_features : int
Number of input channels.
out_features : int
Number of output channels.
"""
def __init__(self, in_features, out_features):
super().__init__()
mid_features = in_features // 4
self.fc1 = nn.Linear(in_features=in_features, out_features=mid_features
)
self.activ = nn.ReLU(inplace=True)
self.fc2 = nn.Linear(in_features=mid_features, out_features=
out_features)
def forward(self, x):
x = self.fc1(x)
x = self.activ(x)
x = self.fc2(x)
return x
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4, 4]
|
AlexConv | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/cc/ccc3jyqbz5rjanlhmaez5ahwrmffxjcm5mibxaw5efspqo45e4re.py
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.convolution]
# Source node to ATen node mapping:
# x => convolution
# Graph fragment:
# %convolution : [num_users=2] = call_function[target=torch.ops.aten.convolution.default](args = (%primals_3, %primals_1, %primals_2, [1, 1], [4, 4], [1, 1], False, [0, 0], 1), kwargs = {})
triton_poi_fused_convolution_0 = async_compile.triton('triton_poi_fused_convolution_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[2048],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_0(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 1296
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = (xindex // 81) % 4
tmp0 = tl.load(in_out_ptr0 + (x3), xmask)
tmp1 = tl.load(in_ptr0 + (x1), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + (x3), tmp2, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/pc/cpch32gty44h5lucgufpgnx36uesj7ujzv57lrbv6jhdocetub5e.py
# Topologically Sorted Source Nodes: [x_2], Original ATen: [aten.constant_pad_nd]
# Source node to ATen node mapping:
# x_2 => constant_pad_nd
# Graph fragment:
# %constant_pad_nd : [num_users=2] = call_function[target=torch.ops.aten.constant_pad_nd.default](args = (%view, [0, 0, 0, 0, 2, 2], 0.0), kwargs = {})
triton_poi_fused_constant_pad_nd_1 = async_compile.triton('triton_poi_fused_constant_pad_nd_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[4096],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_constant_pad_nd_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_constant_pad_nd_1(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 2592
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = (xindex // 81) % 8
x2 = (xindex // 648)
x3 = xindex % 648
x4 = xindex
tmp0 = (-2) + x1
tmp1 = tl.full([1], 0, tl.int64)
tmp2 = tmp0 >= tmp1
tmp3 = tl.full([1], 4, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = tmp2 & tmp4
tmp6 = tl.load(in_ptr0 + ((-162) + x3 + (324*x2)), tmp5 & xmask, other=0.0)
tmp7 = tl.full([1], 0, tl.int32)
tmp8 = triton_helpers.maximum(tmp7, tmp6)
tmp9 = tmp8 * tmp8
tmp10 = tl.full(tmp9.shape, 0.0, tmp9.dtype)
tmp11 = tl.where(tmp5, tmp9, tmp10)
tl.store(out_ptr0 + (x4), tmp11, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/27/c27wmixqxfyefdidpse275doschj5umpqz2ohwf6sinnekud4xnj.py
# Topologically Sorted Source Nodes: [x_1, x_2], Original ATen: [aten.relu, aten.avg_pool3d, aten.mul, aten.add, aten.pow, aten.div]
# Source node to ATen node mapping:
# x_1 => relu
# x_2 => add, avg_pool3d, div, mul_1, pow_1
# Graph fragment:
# %relu : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%convolution,), kwargs = {})
# %avg_pool3d : [num_users=2] = call_function[target=torch.ops.aten.avg_pool3d.default](args = (%constant_pad_nd, [5, 1, 1], [1, 1, 1]), kwargs = {})
# %mul_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%squeeze, 0.0001), kwargs = {})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul_1, 2.0), kwargs = {})
# %pow_1 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%add, 0.75), kwargs = {})
# %div : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%relu, %pow_1), kwargs = {})
triton_poi_fused_add_avg_pool3d_div_mul_pow_relu_2 = async_compile.triton('triton_poi_fused_add_avg_pool3d_div_mul_pow_relu_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[2048],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_avg_pool3d_div_mul_pow_relu_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 6, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_avg_pool3d_div_mul_pow_relu_2(in_ptr0, in_ptr1, out_ptr0, out_ptr1, xnumel, XBLOCK : tl.constexpr):
xnumel = 1296
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 324
x1 = (xindex // 324)
x2 = xindex
tmp0 = tl.load(in_ptr0 + (x0 + (648*x1)), xmask)
tmp1 = tl.load(in_ptr0 + (81 + x0 + (648*x1)), xmask)
tmp3 = tl.load(in_ptr0 + (162 + x0 + (648*x1)), xmask)
tmp5 = tl.load(in_ptr0 + (243 + x0 + (648*x1)), xmask)
tmp7 = tl.load(in_ptr0 + (324 + x0 + (648*x1)), xmask)
tmp11 = tl.load(in_ptr1 + (x2), xmask)
tmp2 = tmp1 + tmp0
tmp4 = tmp3 + tmp2
tmp6 = tmp5 + tmp4
tmp8 = tmp7 + tmp6
tmp9 = 0.2
tmp10 = tmp8 * tmp9
tmp12 = tl.full([1], 0, tl.int32)
tmp13 = triton_helpers.maximum(tmp12, tmp11)
tmp14 = 0.0001
tmp15 = tmp10 * tmp14
tmp16 = 2.0
tmp17 = tmp15 + tmp16
tmp18 = 0.75
tmp19 = libdevice.pow(tmp17, tmp18)
tmp20 = tmp13 / tmp19
tl.store(out_ptr0 + (x2), tmp10, xmask)
tl.store(out_ptr1 + (x2), tmp20, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (4, ), (1, ))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.convolution]
buf0 = extern_kernels.convolution(primals_3, primals_1, stride=(1, 1), padding=(4, 4), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf0, (4, 4, 9, 9), (324, 81, 9, 1))
buf1 = buf0; del buf0 # reuse
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.convolution]
stream0 = get_raw_stream(0)
triton_poi_fused_convolution_0.run(buf1, primals_2, 1296, grid=grid(1296), stream=stream0)
del primals_2
buf2 = empty_strided_cuda((4, 1, 8, 9, 9), (648, 648, 81, 9, 1), torch.float32)
# Topologically Sorted Source Nodes: [x_2], Original ATen: [aten.constant_pad_nd]
triton_poi_fused_constant_pad_nd_1.run(buf1, buf2, 2592, grid=grid(2592), stream=stream0)
buf3 = empty_strided_cuda((4, 1, 4, 9, 9), (324, 324, 81, 9, 1), torch.float32)
buf4 = empty_strided_cuda((4, 4, 9, 9), (324, 81, 9, 1), torch.float32)
# Topologically Sorted Source Nodes: [x_1, x_2], Original ATen: [aten.relu, aten.avg_pool3d, aten.mul, aten.add, aten.pow, aten.div]
triton_poi_fused_add_avg_pool3d_div_mul_pow_relu_2.run(buf2, buf1, buf3, buf4, 1296, grid=grid(1296), stream=stream0)
return (buf4, primals_1, primals_3, buf1, buf2, buf3, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.nn.functional as F
from inspect import isfunction
def get_activation_layer(activation):
"""
Create activation layer from string/function.
Parameters:
----------
activation : function, or str, or nn.Module
Activation function or name of activation function.
Returns
-------
nn.Module
Activation layer.
"""
assert activation is not None
if isfunction(activation):
return activation()
elif isinstance(activation, str):
if activation == 'relu':
return nn.ReLU(inplace=True)
elif activation == 'relu6':
return nn.ReLU6(inplace=True)
elif activation == 'swish':
return Swish()
elif activation == 'hswish':
return HSwish(inplace=True)
elif activation == 'sigmoid':
return nn.Sigmoid()
elif activation == 'hsigmoid':
return HSigmoid()
else:
raise NotImplementedError()
else:
assert isinstance(activation, nn.Module)
return activation
class HSigmoid(nn.Module):
"""
Approximated sigmoid function, so-called hard-version of sigmoid from 'Searching for MobileNetV3,'
https://arxiv.org/abs/1905.02244.
"""
def forward(self, x):
return F.relu6(x + 3.0, inplace=True) / 6.0
class HSwish(nn.Module):
"""
H-Swish activation function from 'Searching for MobileNetV3,' https://arxiv.org/abs/1905.02244.
Parameters:
----------
inplace : bool
Whether to use inplace version of the module.
"""
def __init__(self, inplace=False):
super(HSwish, self).__init__()
self.inplace = inplace
def forward(self, x):
return x * F.relu6(x + 3.0, inplace=self.inplace) / 6.0
class Swish(nn.Module):
"""
Swish activation function from 'Searching for Activation Functions,' https://arxiv.org/abs/1710.05941.
"""
def forward(self, x):
return x * torch.sigmoid(x)
class ConvBlock(nn.Module):
"""
Standard convolution block with Batch normalization and activation.
Parameters:
----------
in_channels : int
Number of input channels.
out_channels : int
Number of output channels.
kernel_size : int or tuple/list of 2 int
Convolution window size.
stride : int or tuple/list of 2 int
Strides of the convolution.
padding : int or tuple/list of 2 int
Padding value for convolution layer.
dilation : int or tuple/list of 2 int, default 1
Dilation value for convolution layer.
groups : int, default 1
Number of groups.
bias : bool, default False
Whether the layer uses a bias vector.
use_bn : bool, default True
Whether to use BatchNorm layer.
bn_eps : float, default 1e-5
Small float added to variance in Batch norm.
activation : function or str or None, default nn.ReLU(inplace=True)
Activation function or name of activation function.
"""
def __init__(self, in_channels, out_channels, kernel_size, stride,
padding, dilation=1, groups=1, bias=False, use_bn=True, bn_eps=
1e-05, activation=lambda : nn.ReLU(inplace=True)):
super(ConvBlock, self).__init__()
self.activate = activation is not None
self.use_bn = use_bn
self.conv = nn.Conv2d(in_channels=in_channels, out_channels=
out_channels, kernel_size=kernel_size, stride=stride, padding=
padding, dilation=dilation, groups=groups, bias=bias)
if self.use_bn:
self.bn = nn.BatchNorm2d(num_features=out_channels, eps=bn_eps)
if self.activate:
self.activ = get_activation_layer(activation)
def forward(self, x):
x = self.conv(x)
if self.use_bn:
x = self.bn(x)
if self.activate:
x = self.activ(x)
return x
class AlexConv(ConvBlock):
"""
AlexNet specific convolution block.
Parameters:
----------
in_channels : int
Number of input channels.
out_channels : int
Number of output channels.
kernel_size : int or tuple/list of 2 int
Convolution window size.
stride : int or tuple/list of 2 int
Strides of the convolution.
padding : int or tuple/list of 2 int
Padding value for convolution layer.
use_lrn : bool
Whether to use LRN layer.
"""
def __init__(self, in_channels, out_channels, kernel_size, stride,
padding, use_lrn):
super(AlexConv, self).__init__(in_channels=in_channels,
out_channels=out_channels, kernel_size=kernel_size, stride=
stride, padding=padding, bias=True, use_bn=False)
self.use_lrn = use_lrn
def forward(self, x):
x = super(AlexConv, self).forward(x)
if self.use_lrn:
x = F.local_response_norm(x, size=5, k=2.0)
return x
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'in_channels': 4, 'out_channels': 4, 'kernel_size': 4,
'stride': 1, 'padding': 4, 'use_lrn': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import libdevice
import torch.nn as nn
import torch.nn.functional as F
from inspect import isfunction
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_poi_fused_convolution_0(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl
.constexpr):
xnumel = 1296
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = xindex // 81 % 4
tmp0 = tl.load(in_out_ptr0 + x3, xmask)
tmp1 = tl.load(in_ptr0 + x1, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + x3, tmp2, xmask)
@triton.jit
def triton_poi_fused_constant_pad_nd_1(in_ptr0, out_ptr0, xnumel, XBLOCK:
tl.constexpr):
xnumel = 2592
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = xindex // 81 % 8
x2 = xindex // 648
x3 = xindex % 648
x4 = xindex
tmp0 = -2 + x1
tmp1 = tl.full([1], 0, tl.int64)
tmp2 = tmp0 >= tmp1
tmp3 = tl.full([1], 4, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = tmp2 & tmp4
tmp6 = tl.load(in_ptr0 + (-162 + x3 + 324 * x2), tmp5 & xmask, other=0.0)
tmp7 = tl.full([1], 0, tl.int32)
tmp8 = triton_helpers.maximum(tmp7, tmp6)
tmp9 = tmp8 * tmp8
tmp10 = tl.full(tmp9.shape, 0.0, tmp9.dtype)
tmp11 = tl.where(tmp5, tmp9, tmp10)
tl.store(out_ptr0 + x4, tmp11, xmask)
@triton.jit
def triton_poi_fused_add_avg_pool3d_div_mul_pow_relu_2(in_ptr0, in_ptr1,
out_ptr0, out_ptr1, xnumel, XBLOCK: tl.constexpr):
xnumel = 1296
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 324
x1 = xindex // 324
x2 = xindex
tmp0 = tl.load(in_ptr0 + (x0 + 648 * x1), xmask)
tmp1 = tl.load(in_ptr0 + (81 + x0 + 648 * x1), xmask)
tmp3 = tl.load(in_ptr0 + (162 + x0 + 648 * x1), xmask)
tmp5 = tl.load(in_ptr0 + (243 + x0 + 648 * x1), xmask)
tmp7 = tl.load(in_ptr0 + (324 + x0 + 648 * x1), xmask)
tmp11 = tl.load(in_ptr1 + x2, xmask)
tmp2 = tmp1 + tmp0
tmp4 = tmp3 + tmp2
tmp6 = tmp5 + tmp4
tmp8 = tmp7 + tmp6
tmp9 = 0.2
tmp10 = tmp8 * tmp9
tmp12 = tl.full([1], 0, tl.int32)
tmp13 = triton_helpers.maximum(tmp12, tmp11)
tmp14 = 0.0001
tmp15 = tmp10 * tmp14
tmp16 = 2.0
tmp17 = tmp15 + tmp16
tmp18 = 0.75
tmp19 = libdevice.pow(tmp17, tmp18)
tmp20 = tmp13 / tmp19
tl.store(out_ptr0 + x2, tmp10, xmask)
tl.store(out_ptr1 + x2, tmp20, xmask)
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (4,), (1,))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = extern_kernels.convolution(primals_3, primals_1, stride=(1,
1), padding=(4, 4), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf0, (4, 4, 9, 9), (324, 81, 9, 1))
buf1 = buf0
del buf0
get_raw_stream(0)
triton_poi_fused_convolution_0[grid(1296)](buf1, primals_2, 1296,
XBLOCK=256, num_warps=4, num_stages=1)
del primals_2
buf2 = empty_strided_cuda((4, 1, 8, 9, 9), (648, 648, 81, 9, 1),
torch.float32)
triton_poi_fused_constant_pad_nd_1[grid(2592)](buf1, buf2, 2592,
XBLOCK=128, num_warps=4, num_stages=1)
buf3 = empty_strided_cuda((4, 1, 4, 9, 9), (324, 324, 81, 9, 1),
torch.float32)
buf4 = empty_strided_cuda((4, 4, 9, 9), (324, 81, 9, 1), torch.float32)
triton_poi_fused_add_avg_pool3d_div_mul_pow_relu_2[grid(1296)](buf2,
buf1, buf3, buf4, 1296, XBLOCK=128, num_warps=4, num_stages=1)
return buf4, primals_1, primals_3, buf1, buf2, buf3
def get_activation_layer(activation):
"""
Create activation layer from string/function.
Parameters:
----------
activation : function, or str, or nn.Module
Activation function or name of activation function.
Returns
-------
nn.Module
Activation layer.
"""
assert activation is not None
if isfunction(activation):
return activation()
elif isinstance(activation, str):
if activation == 'relu':
return nn.ReLU(inplace=True)
elif activation == 'relu6':
return nn.ReLU6(inplace=True)
elif activation == 'swish':
return Swish()
elif activation == 'hswish':
return HSwish(inplace=True)
elif activation == 'sigmoid':
return nn.Sigmoid()
elif activation == 'hsigmoid':
return HSigmoid()
else:
raise NotImplementedError()
else:
assert isinstance(activation, nn.Module)
return activation
class HSigmoid(nn.Module):
"""
Approximated sigmoid function, so-called hard-version of sigmoid from 'Searching for MobileNetV3,'
https://arxiv.org/abs/1905.02244.
"""
def forward(self, x):
return F.relu6(x + 3.0, inplace=True) / 6.0
class HSwish(nn.Module):
"""
H-Swish activation function from 'Searching for MobileNetV3,' https://arxiv.org/abs/1905.02244.
Parameters:
----------
inplace : bool
Whether to use inplace version of the module.
"""
def __init__(self, inplace=False):
super(HSwish, self).__init__()
self.inplace = inplace
def forward(self, x):
return x * F.relu6(x + 3.0, inplace=self.inplace) / 6.0
class Swish(nn.Module):
"""
Swish activation function from 'Searching for Activation Functions,' https://arxiv.org/abs/1710.05941.
"""
def forward(self, x):
return x * torch.sigmoid(x)
class ConvBlock(nn.Module):
"""
Standard convolution block with Batch normalization and activation.
Parameters:
----------
in_channels : int
Number of input channels.
out_channels : int
Number of output channels.
kernel_size : int or tuple/list of 2 int
Convolution window size.
stride : int or tuple/list of 2 int
Strides of the convolution.
padding : int or tuple/list of 2 int
Padding value for convolution layer.
dilation : int or tuple/list of 2 int, default 1
Dilation value for convolution layer.
groups : int, default 1
Number of groups.
bias : bool, default False
Whether the layer uses a bias vector.
use_bn : bool, default True
Whether to use BatchNorm layer.
bn_eps : float, default 1e-5
Small float added to variance in Batch norm.
activation : function or str or None, default nn.ReLU(inplace=True)
Activation function or name of activation function.
"""
def __init__(self, in_channels, out_channels, kernel_size, stride,
padding, dilation=1, groups=1, bias=False, use_bn=True, bn_eps=
1e-05, activation=lambda : nn.ReLU(inplace=True)):
super(ConvBlock, self).__init__()
self.activate = activation is not None
self.use_bn = use_bn
self.conv = nn.Conv2d(in_channels=in_channels, out_channels=
out_channels, kernel_size=kernel_size, stride=stride, padding=
padding, dilation=dilation, groups=groups, bias=bias)
if self.use_bn:
self.bn = nn.BatchNorm2d(num_features=out_channels, eps=bn_eps)
if self.activate:
self.activ = get_activation_layer(activation)
def forward(self, x):
x = self.conv(x)
if self.use_bn:
x = self.bn(x)
if self.activate:
x = self.activ(x)
return x
class AlexConvNew(ConvBlock):
"""
AlexNet specific convolution block.
Parameters:
----------
in_channels : int
Number of input channels.
out_channels : int
Number of output channels.
kernel_size : int or tuple/list of 2 int
Convolution window size.
stride : int or tuple/list of 2 int
Strides of the convolution.
padding : int or tuple/list of 2 int
Padding value for convolution layer.
use_lrn : bool
Whether to use LRN layer.
"""
def __init__(self, in_channels, out_channels, kernel_size, stride,
padding, use_lrn):
super(AlexConvNew, self).__init__(in_channels=in_channels,
out_channels=out_channels, kernel_size=kernel_size, stride=
stride, padding=padding, bias=True, use_bn=False)
self.use_lrn = use_lrn
def forward(self, input_0):
primals_1 = self.conv.weight
primals_2 = self.conv.bias
primals_3 = input_0
output = call([primals_1, primals_2, primals_3])
return output[0]
| iofthetiger/pkuad | AlexConv | false | 6,899 | [
"Apache-2.0"
] | 1 | 07496d108c614c84be028f344830becc9cac8fe5 | https://github.com/iofthetiger/pkuad/tree/07496d108c614c84be028f344830becc9cac8fe5 | import torch
import torch.nn as nn
import torch.nn.functional as F
from inspect import isfunction
def get_activation_layer(activation):
"""
Create activation layer from string/function.
Parameters:
----------
activation : function, or str, or nn.Module
Activation function or name of activation function.
Returns
-------
nn.Module
Activation layer.
"""
assert activation is not None
if isfunction(activation):
return activation()
elif isinstance(activation, str):
if activation == 'relu':
return nn.ReLU(inplace=True)
elif activation == 'relu6':
return nn.ReLU6(inplace=True)
elif activation == 'swish':
return Swish()
elif activation == 'hswish':
return HSwish(inplace=True)
elif activation == 'sigmoid':
return nn.Sigmoid()
elif activation == 'hsigmoid':
return HSigmoid()
else:
raise NotImplementedError()
else:
assert isinstance(activation, nn.Module)
return activation
class HSigmoid(nn.Module):
"""
Approximated sigmoid function, so-called hard-version of sigmoid from 'Searching for MobileNetV3,'
https://arxiv.org/abs/1905.02244.
"""
def forward(self, x):
return F.relu6(x + 3.0, inplace=True) / 6.0
class HSwish(nn.Module):
"""
H-Swish activation function from 'Searching for MobileNetV3,' https://arxiv.org/abs/1905.02244.
Parameters:
----------
inplace : bool
Whether to use inplace version of the module.
"""
def __init__(self, inplace=False):
super().__init__()
self.inplace = inplace
def forward(self, x):
return x * F.relu6(x + 3.0, inplace=self.inplace) / 6.0
class Swish(nn.Module):
"""
Swish activation function from 'Searching for Activation Functions,' https://arxiv.org/abs/1710.05941.
"""
def forward(self, x):
return x * torch.sigmoid(x)
class ConvBlock(nn.Module):
"""
Standard convolution block with Batch normalization and activation.
Parameters:
----------
in_channels : int
Number of input channels.
out_channels : int
Number of output channels.
kernel_size : int or tuple/list of 2 int
Convolution window size.
stride : int or tuple/list of 2 int
Strides of the convolution.
padding : int or tuple/list of 2 int
Padding value for convolution layer.
dilation : int or tuple/list of 2 int, default 1
Dilation value for convolution layer.
groups : int, default 1
Number of groups.
bias : bool, default False
Whether the layer uses a bias vector.
use_bn : bool, default True
Whether to use BatchNorm layer.
bn_eps : float, default 1e-5
Small float added to variance in Batch norm.
activation : function or str or None, default nn.ReLU(inplace=True)
Activation function or name of activation function.
"""
def __init__(self, in_channels, out_channels, kernel_size, stride,
padding, dilation=1, groups=1, bias=False, use_bn=True, bn_eps=
1e-05, activation=lambda : nn.ReLU(inplace=True)):
super().__init__()
self.activate = activation is not None
self.use_bn = use_bn
self.conv = nn.Conv2d(in_channels=in_channels, out_channels=
out_channels, kernel_size=kernel_size, stride=stride, padding=
padding, dilation=dilation, groups=groups, bias=bias)
if self.use_bn:
self.bn = nn.BatchNorm2d(num_features=out_channels, eps=bn_eps)
if self.activate:
self.activ = get_activation_layer(activation)
def forward(self, x):
x = self.conv(x)
if self.use_bn:
x = self.bn(x)
if self.activate:
x = self.activ(x)
return x
class Model(ConvBlock):
"""
AlexNet specific convolution block.
# ... truncated (>4000 chars) for memory efficiency |
Attention | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/um/cum65j23qchrjf5dndblqgbw6zomhgwfj2obfidtgy7b5j3zwklm.py
# Topologically Sorted Source Nodes: [posts_attention_weights], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# posts_attention_weights => amax, exp, sub
# Graph fragment:
# %amax : [num_users=1] = call_function[target=torch.ops.aten.amax.default](args = (%permute_1, [-1], True), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%permute_1, %amax), kwargs = {})
# %exp : [num_users=2] = call_function[target=torch.ops.aten.exp.default](args = (%sub,), kwargs = {})
triton_poi_fused__softmax_0 = async_compile.triton('triton_poi_fused__softmax_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_0(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp3 = triton_helpers.maximum(tmp1, tmp2)
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp7 = triton_helpers.maximum(tmp5, tmp6)
tmp8 = tmp0 - tmp7
tmp9 = tl_math.exp(tmp8)
tl.store(out_ptr0 + (x2), tmp9, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/wk/cwk2wao7opapqbjj7klnqrd6tgist3ts3nc5veryzhzstwpx7d4l.py
# Topologically Sorted Source Nodes: [posts_attention_weights], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# posts_attention_weights => div, sum_1
# Graph fragment:
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%exp, [-1], True), kwargs = {})
# %div : [num_users=2] = call_function[target=torch.ops.aten.div.Tensor](args = (%exp, %sum_1), kwargs = {})
triton_poi_fused__softmax_1 = async_compile.triton('triton_poi_fused__softmax_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_1(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + (x2), tmp8, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (1, 4), (4, 1))
assert_size_stride(primals_2, (1, ), (1, ))
assert_size_stride(primals_3, (4, 4, 4), (16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf1 = empty_strided_cuda((16, 1), (1, 1), torch.float32)
# Topologically Sorted Source Nodes: [posts_attention_values], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_2, reinterpret_tensor(primals_3, (16, 4), (4, 1), 0), reinterpret_tensor(primals_1, (4, 1), (1, 4), 0), alpha=1, beta=1, out=buf1)
del primals_1
del primals_2
buf2 = empty_strided_cuda((4, 1, 4), (4, 16, 1), torch.float32)
# Topologically Sorted Source Nodes: [posts_attention_weights], Original ATen: [aten._softmax]
stream0 = get_raw_stream(0)
triton_poi_fused__softmax_0.run(buf1, buf2, 16, grid=grid(16), stream=stream0)
buf3 = reinterpret_tensor(buf1, (4, 1, 4), (4, 4, 1), 0); del buf1 # reuse
# Topologically Sorted Source Nodes: [posts_attention_weights], Original ATen: [aten._softmax]
triton_poi_fused__softmax_1.run(buf2, buf3, 16, grid=grid(16), stream=stream0)
buf4 = reinterpret_tensor(buf2, (4, 1, 4), (4, 4, 1), 0); del buf2 # reuse
# Topologically Sorted Source Nodes: [matmul], Original ATen: [aten.bmm]
extern_kernels.bmm(buf3, primals_3, out=buf4)
return (reinterpret_tensor(buf4, (4, 4), (4, 1), 0), buf3, primals_3, buf3, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((1, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((1, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn.functional as F
class Attention(torch.nn.Module):
"""Scaled dot product attention."""
def __init__(self, hidden_dim, **kwargs):
super(Attention, self).__init__(**kwargs)
self.projection_layer = torch.nn.Linear(hidden_dim, 1)
def forward(self, atten_post):
posts_attention_values = self.projection_layer(atten_post)
posts_attention_weights = F.softmax(posts_attention_values.permute(
0, 2, 1), dim=-1)
del posts_attention_values
torch.cuda.empty_cache()
self_atten_output_post = torch.matmul(posts_attention_weights,
atten_post).squeeze(dim=1)
return self_atten_output_post, posts_attention_weights
def get_inputs():
return [torch.rand([4, 4, 4])]
def get_init_inputs():
return [[], {'hidden_dim': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import math as tl_math
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused__softmax_0(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp3 = triton_helpers.maximum(tmp1, tmp2)
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp7 = triton_helpers.maximum(tmp5, tmp6)
tmp8 = tmp0 - tmp7
tmp9 = tl_math.exp(tmp8)
tl.store(out_ptr0 + x2, tmp9, xmask)
@triton.jit
def triton_poi_fused__softmax_1(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + x2, tmp8, xmask)
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (1, 4), (4, 1))
assert_size_stride(primals_2, (1,), (1,))
assert_size_stride(primals_3, (4, 4, 4), (16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf1 = empty_strided_cuda((16, 1), (1, 1), torch.float32)
extern_kernels.addmm(primals_2, reinterpret_tensor(primals_3, (16,
4), (4, 1), 0), reinterpret_tensor(primals_1, (4, 1), (1, 4), 0
), alpha=1, beta=1, out=buf1)
del primals_1
del primals_2
buf2 = empty_strided_cuda((4, 1, 4), (4, 16, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused__softmax_0[grid(16)](buf1, buf2, 16, XBLOCK=16,
num_warps=1, num_stages=1)
buf3 = reinterpret_tensor(buf1, (4, 1, 4), (4, 4, 1), 0)
del buf1
triton_poi_fused__softmax_1[grid(16)](buf2, buf3, 16, XBLOCK=16,
num_warps=1, num_stages=1)
buf4 = reinterpret_tensor(buf2, (4, 1, 4), (4, 4, 1), 0)
del buf2
extern_kernels.bmm(buf3, primals_3, out=buf4)
return reinterpret_tensor(buf4, (4, 4), (4, 1), 0), buf3, primals_3, buf3
class AttentionNew(torch.nn.Module):
"""Scaled dot product attention."""
def __init__(self, hidden_dim, **kwargs):
super(AttentionNew, self).__init__(**kwargs)
self.projection_layer = torch.nn.Linear(hidden_dim, 1)
def forward(self, input_0):
primals_1 = self.projection_layer.weight
primals_2 = self.projection_layer.bias
primals_3 = input_0
output = call([primals_1, primals_2, primals_3])
return output[0], output[1]
| isspek/Cross-Lingual-Cyberbullying | Attention | false | 6,900 | [
"MIT"
] | 1 | 710c136b9233f0be87af72e43e25722e73158c52 | https://github.com/isspek/Cross-Lingual-Cyberbullying/tree/710c136b9233f0be87af72e43e25722e73158c52 | import torch
import torch.nn.functional as F
class Model(torch.nn.Module):
"""Scaled dot product attention."""
def __init__(self, hidden_dim, **kwargs):
super().__init__(**kwargs)
self.projection_layer = torch.nn.Linear(hidden_dim, 1)
def forward(self, atten_post):
posts_attention_values = self.projection_layer(atten_post)
posts_attention_weights = F.softmax(posts_attention_values.permute(
0, 2, 1), dim=-1)
del posts_attention_values
torch.cuda.empty_cache()
self_atten_output_post = torch.matmul(posts_attention_weights,
atten_post).squeeze(dim=1)
return self_atten_output_post, posts_attention_weights
def get_inputs():
return [torch.rand([4, 4, 4])]
def get_init_inputs():
return [4]
|
MobileNetV3Classifier | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/jj/cjjcpa4jfom3kmx4ufnxtda3bmq466cpemkegyhzep2ymmlsg35l.py
# Topologically Sorted Source Nodes: [add, relu6, mul, x_1], Original ATen: [aten.add, aten.hardtanh, aten.mul, aten.div]
# Source node to ATen node mapping:
# add => add
# mul => mul
# relu6 => clamp_max, clamp_min
# x_1 => div
# Graph fragment:
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%convolution, 3.0), kwargs = {})
# %clamp_min : [num_users=1] = call_function[target=torch.ops.aten.clamp_min.default](args = (%add, 0), kwargs = {})
# %clamp_max : [num_users=1] = call_function[target=torch.ops.aten.clamp_max.default](args = (%clamp_min, 6), kwargs = {})
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%convolution, %clamp_max), kwargs = {})
# %div : [num_users=2] = call_function[target=torch.ops.aten.div.Tensor](args = (%mul, 6.0), kwargs = {})
triton_poi_fused_add_div_hardtanh_mul_0 = async_compile.triton('triton_poi_fused_add_div_hardtanh_mul_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_div_hardtanh_mul_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_div_hardtanh_mul_0(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (x0), xmask)
tmp1 = 3.0
tmp2 = tmp0 + tmp1
tmp3 = 0.0
tmp4 = triton_helpers.maximum(tmp2, tmp3)
tmp5 = 6.0
tmp6 = triton_helpers.minimum(tmp4, tmp5)
tmp7 = tmp0 * tmp6
tmp8 = 0.16666666666666666
tmp9 = tmp7 * tmp8
tl.store(out_ptr0 + (x0), tmp9, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/32/c32v7egt4mupqssam3gmac2qgv3ujprjybthsgweflmot256qqw7.py
# Topologically Sorted Source Nodes: [x_3], Original ATen: [aten.convolution]
# Source node to ATen node mapping:
# x_3 => convolution_1
# Graph fragment:
# %convolution_1 : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%div, %primals_3, %primals_4, [1, 1], [0, 0], [1, 1], False, [0, 0], 1), kwargs = {})
triton_poi_fused_convolution_1 = async_compile.triton('triton_poi_fused_convolution_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_1', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_1(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = (xindex // 16) % 4
tmp0 = tl.load(in_out_ptr0 + (x3), xmask)
tmp1 = tl.load(in_ptr0 + (x1), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + (x3), tmp2, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 1, 1), (4, 1, 1, 1))
assert_size_stride(primals_2, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_3, (4, 4, 1, 1), (4, 1, 1, 1))
assert_size_stride(primals_4, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.convolution]
buf0 = extern_kernels.convolution(primals_2, primals_1, stride=(1, 1), padding=(0, 0), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf0, (4, 4, 4, 4), (64, 16, 4, 1))
buf1 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [add, relu6, mul, x_1], Original ATen: [aten.add, aten.hardtanh, aten.mul, aten.div]
stream0 = get_raw_stream(0)
triton_poi_fused_add_div_hardtanh_mul_0.run(buf0, buf1, 256, grid=grid(256), stream=stream0)
# Topologically Sorted Source Nodes: [x_3], Original ATen: [aten.convolution]
buf2 = extern_kernels.convolution(buf1, primals_3, stride=(1, 1), padding=(0, 0), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf2, (4, 4, 4, 4), (64, 16, 4, 1))
buf3 = buf2; del buf2 # reuse
# Topologically Sorted Source Nodes: [x_3], Original ATen: [aten.convolution]
triton_poi_fused_convolution_1.run(buf3, primals_4, 256, grid=grid(256), stream=stream0)
del primals_4
return (buf3, primals_1, primals_2, primals_3, buf0, buf1, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 1, 1), (4, 1, 1, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4, 1, 1), (4, 1, 1, 1), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.nn.functional as F
def conv1x1(in_channels, out_channels, stride=1, groups=1, bias=False):
"""
Convolution 1x1 layer.
Parameters:
----------
in_channels : int
Number of input channels.
out_channels : int
Number of output channels.
stride : int or tuple/list of 2 int, default 1
Strides of the convolution.
groups : int, default 1
Number of groups.
bias : bool, default False
Whether the layer uses a bias vector.
"""
return nn.Conv2d(in_channels=in_channels, out_channels=out_channels,
kernel_size=1, stride=stride, groups=groups, bias=bias)
class HSwish(nn.Module):
"""
H-Swish activation function from 'Searching for MobileNetV3,' https://arxiv.org/abs/1905.02244.
Parameters:
----------
inplace : bool
Whether to use inplace version of the module.
"""
def __init__(self, inplace=False):
super(HSwish, self).__init__()
self.inplace = inplace
def forward(self, x):
return x * F.relu6(x + 3.0, inplace=self.inplace) / 6.0
class MobileNetV3Classifier(nn.Module):
"""
MobileNetV3 classifier.
Parameters:
----------
in_channels : int
Number of input channels.
out_channels : int
Number of output channels.
mid_channels : int
Number of middle channels.
dropout_rate : float
Parameter of Dropout layer. Faction of the input units to drop.
"""
def __init__(self, in_channels, out_channels, mid_channels, dropout_rate):
super(MobileNetV3Classifier, self).__init__()
self.use_dropout = dropout_rate != 0.0
self.conv1 = conv1x1(in_channels=in_channels, out_channels=mid_channels
)
self.activ = HSwish(inplace=True)
if self.use_dropout:
self.dropout = nn.Dropout(p=dropout_rate)
self.conv2 = conv1x1(in_channels=mid_channels, out_channels=
out_channels, bias=True)
def forward(self, x):
x = self.conv1(x)
x = self.activ(x)
if self.use_dropout:
x = self.dropout(x)
x = self.conv2(x)
return x
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'in_channels': 4, 'out_channels': 4, 'mid_channels': 4,
'dropout_rate': 0.5}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
import torch.nn as nn
import torch.nn.functional as F
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_poi_fused_add_div_hardtanh_mul_0(in_ptr0, out_ptr0, xnumel,
XBLOCK: tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + x0, xmask)
tmp1 = 3.0
tmp2 = tmp0 + tmp1
tmp3 = 0.0
tmp4 = triton_helpers.maximum(tmp2, tmp3)
tmp5 = 6.0
tmp6 = triton_helpers.minimum(tmp4, tmp5)
tmp7 = tmp0 * tmp6
tmp8 = 0.16666666666666666
tmp9 = tmp7 * tmp8
tl.store(out_ptr0 + x0, tmp9, xmask)
@triton.jit
def triton_poi_fused_convolution_1(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl
.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = xindex // 16 % 4
tmp0 = tl.load(in_out_ptr0 + x3, xmask)
tmp1 = tl.load(in_ptr0 + x1, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + x3, tmp2, xmask)
def call(args):
primals_1, primals_2, primals_3, primals_4 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 1, 1), (4, 1, 1, 1))
assert_size_stride(primals_2, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_3, (4, 4, 1, 1), (4, 1, 1, 1))
assert_size_stride(primals_4, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = extern_kernels.convolution(primals_2, primals_1, stride=(1,
1), padding=(0, 0), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf0, (4, 4, 4, 4), (64, 16, 4, 1))
buf1 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_add_div_hardtanh_mul_0[grid(256)](buf0, buf1, 256,
XBLOCK=128, num_warps=4, num_stages=1)
buf2 = extern_kernels.convolution(buf1, primals_3, stride=(1, 1),
padding=(0, 0), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf2, (4, 4, 4, 4), (64, 16, 4, 1))
buf3 = buf2
del buf2
triton_poi_fused_convolution_1[grid(256)](buf3, primals_4, 256,
XBLOCK=256, num_warps=4, num_stages=1)
del primals_4
return buf3, primals_1, primals_2, primals_3, buf0, buf1
def conv1x1(in_channels, out_channels, stride=1, groups=1, bias=False):
"""
Convolution 1x1 layer.
Parameters:
----------
in_channels : int
Number of input channels.
out_channels : int
Number of output channels.
stride : int or tuple/list of 2 int, default 1
Strides of the convolution.
groups : int, default 1
Number of groups.
bias : bool, default False
Whether the layer uses a bias vector.
"""
return nn.Conv2d(in_channels=in_channels, out_channels=out_channels,
kernel_size=1, stride=stride, groups=groups, bias=bias)
class HSwish(nn.Module):
"""
H-Swish activation function from 'Searching for MobileNetV3,' https://arxiv.org/abs/1905.02244.
Parameters:
----------
inplace : bool
Whether to use inplace version of the module.
"""
def __init__(self, inplace=False):
super(HSwish, self).__init__()
self.inplace = inplace
def forward(self, x):
return x * F.relu6(x + 3.0, inplace=self.inplace) / 6.0
class MobileNetV3ClassifierNew(nn.Module):
"""
MobileNetV3 classifier.
Parameters:
----------
in_channels : int
Number of input channels.
out_channels : int
Number of output channels.
mid_channels : int
Number of middle channels.
dropout_rate : float
Parameter of Dropout layer. Faction of the input units to drop.
"""
def __init__(self, in_channels, out_channels, mid_channels, dropout_rate):
super(MobileNetV3ClassifierNew, self).__init__()
self.use_dropout = dropout_rate != 0.0
self.conv1 = conv1x1(in_channels=in_channels, out_channels=mid_channels
)
self.activ = HSwish(inplace=True)
if self.use_dropout:
self.dropout = nn.Dropout(p=dropout_rate)
self.conv2 = conv1x1(in_channels=mid_channels, out_channels=
out_channels, bias=True)
def forward(self, input_0):
primals_1 = self.conv1.weight
primals_3 = self.conv2.weight
primals_4 = self.conv2.bias
primals_2 = input_0
output = call([primals_1, primals_2, primals_3, primals_4])
return output[0]
| iofthetiger/pkuad | MobileNetV3Classifier | false | 6,901 | [
"Apache-2.0"
] | 1 | 07496d108c614c84be028f344830becc9cac8fe5 | https://github.com/iofthetiger/pkuad/tree/07496d108c614c84be028f344830becc9cac8fe5 | import torch
import torch.nn as nn
import torch.nn.functional as F
def conv1x1(in_channels, out_channels, stride=1, groups=1, bias=False):
"""
Convolution 1x1 layer.
Parameters:
----------
in_channels : int
Number of input channels.
out_channels : int
Number of output channels.
stride : int or tuple/list of 2 int, default 1
Strides of the convolution.
groups : int, default 1
Number of groups.
bias : bool, default False
Whether the layer uses a bias vector.
"""
return nn.Conv2d(in_channels=in_channels, out_channels=out_channels,
kernel_size=1, stride=stride, groups=groups, bias=bias)
class HSwish(nn.Module):
"""
H-Swish activation function from 'Searching for MobileNetV3,' https://arxiv.org/abs/1905.02244.
Parameters:
----------
inplace : bool
Whether to use inplace version of the module.
"""
def __init__(self, inplace=False):
super().__init__()
self.inplace = inplace
def forward(self, x):
return x * F.relu6(x + 3.0, inplace=self.inplace) / 6.0
class Model(nn.Module):
"""
MobileNetV3 classifier.
Parameters:
----------
in_channels : int
Number of input channels.
out_channels : int
Number of output channels.
mid_channels : int
Number of middle channels.
dropout_rate : float
Parameter of Dropout layer. Faction of the input units to drop.
"""
def __init__(self, in_channels, out_channels, mid_channels, dropout_rate):
super().__init__()
self.use_dropout = dropout_rate != 0.0
self.conv1 = conv1x1(in_channels=in_channels, out_channels=mid_channels
)
self.activ = HSwish(inplace=True)
if self.use_dropout:
self.dropout = nn.Dropout(p=dropout_rate)
self.conv2 = conv1x1(in_channels=mid_channels, out_channels=
out_channels, bias=True)
def forward(self, x):
x = self.conv1(x)
x = self.activ(x)
if self.use_dropout:
x = self.dropout(x)
x = self.conv2(x)
return x
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'in_channels': 4, 'out_channels': 4, 'mid_channels': 4,
'dropout_rate': 0.5}]
|
SPHead | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/td/ctdybbibnws4d7ukbk3fpn35zkgapxylowdhzwx7vgsllncbdrxa.py
# Topologically Sorted Source Nodes: [x, x_1], Original ATen: [aten.convolution, aten.relu]
# Source node to ATen node mapping:
# x => convolution
# x_1 => relu
# Graph fragment:
# %convolution : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%primals_3, %primals_1, %primals_2, [1, 1], [1, 1], [1, 1], False, [0, 0], 1), kwargs = {})
# %relu : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%convolution,), kwargs = {})
triton_poi_fused_convolution_relu_0 = async_compile.triton('triton_poi_fused_convolution_relu_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_relu_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_relu_0(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = (xindex // 16) % 4
tmp0 = tl.load(in_out_ptr0 + (x3), xmask)
tmp1 = tl.load(in_ptr0 + (x1), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tl.store(in_out_ptr0 + (x3), tmp4, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/32/c32v7egt4mupqssam3gmac2qgv3ujprjybthsgweflmot256qqw7.py
# Topologically Sorted Source Nodes: [x_2], Original ATen: [aten.convolution]
# Source node to ATen node mapping:
# x_2 => convolution_1
# Graph fragment:
# %convolution_1 : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%relu, %primals_4, %primals_5, [1, 1], [0, 0], [1, 1], False, [0, 0], 1), kwargs = {})
triton_poi_fused_convolution_1 = async_compile.triton('triton_poi_fused_convolution_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_1', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_1(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = (xindex // 16) % 4
tmp0 = tl.load(in_out_ptr0 + (x3), xmask)
tmp1 = tl.load(in_ptr0 + (x1), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + (x3), tmp2, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 3, 3), (36, 9, 3, 1))
assert_size_stride(primals_2, (4, ), (1, ))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_4, (4, 4, 1, 1), (4, 1, 1, 1))
assert_size_stride(primals_5, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.convolution]
buf0 = extern_kernels.convolution(primals_3, primals_1, stride=(1, 1), padding=(1, 1), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf0, (4, 4, 4, 4), (64, 16, 4, 1))
buf1 = buf0; del buf0 # reuse
# Topologically Sorted Source Nodes: [x, x_1], Original ATen: [aten.convolution, aten.relu]
stream0 = get_raw_stream(0)
triton_poi_fused_convolution_relu_0.run(buf1, primals_2, 256, grid=grid(256), stream=stream0)
del primals_2
# Topologically Sorted Source Nodes: [x_2], Original ATen: [aten.convolution]
buf2 = extern_kernels.convolution(buf1, primals_4, stride=(1, 1), padding=(0, 0), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf2, (4, 4, 4, 4), (64, 16, 4, 1))
buf3 = buf2; del buf2 # reuse
# Topologically Sorted Source Nodes: [x_2], Original ATen: [aten.convolution]
triton_poi_fused_convolution_1.run(buf3, primals_5, 256, grid=grid(256), stream=stream0)
del primals_5
return (buf3, primals_1, primals_3, primals_4, buf1, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 3, 3), (36, 9, 3, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, 4, 1, 1), (4, 1, 1, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.nn.functional as F
from inspect import isfunction
def conv1x1(in_channels, out_channels, stride=1, groups=1, bias=False):
"""
Convolution 1x1 layer.
Parameters:
----------
in_channels : int
Number of input channels.
out_channels : int
Number of output channels.
stride : int or tuple/list of 2 int, default 1
Strides of the convolution.
groups : int, default 1
Number of groups.
bias : bool, default False
Whether the layer uses a bias vector.
"""
return nn.Conv2d(in_channels=in_channels, out_channels=out_channels,
kernel_size=1, stride=stride, groups=groups, bias=bias)
def get_activation_layer(activation):
"""
Create activation layer from string/function.
Parameters:
----------
activation : function, or str, or nn.Module
Activation function or name of activation function.
Returns
-------
nn.Module
Activation layer.
"""
assert activation is not None
if isfunction(activation):
return activation()
elif isinstance(activation, str):
if activation == 'relu':
return nn.ReLU(inplace=True)
elif activation == 'relu6':
return nn.ReLU6(inplace=True)
elif activation == 'swish':
return Swish()
elif activation == 'hswish':
return HSwish(inplace=True)
elif activation == 'sigmoid':
return nn.Sigmoid()
elif activation == 'hsigmoid':
return HSigmoid()
else:
raise NotImplementedError()
else:
assert isinstance(activation, nn.Module)
return activation
def conv3x3_block(in_channels, out_channels, stride=1, padding=1, dilation=
1, groups=1, bias=False, use_bn=True, bn_eps=1e-05, activation=lambda :
nn.ReLU(inplace=True)):
"""
3x3 version of the standard convolution block.
Parameters:
----------
in_channels : int
Number of input channels.
out_channels : int
Number of output channels.
stride : int or tuple/list of 2 int, default 1
Strides of the convolution.
padding : int or tuple/list of 2 int, default 1
Padding value for convolution layer.
dilation : int or tuple/list of 2 int, default 1
Dilation value for convolution layer.
groups : int, default 1
Number of groups.
bias : bool, default False
Whether the layer uses a bias vector.
use_bn : bool, default True
Whether to use BatchNorm layer.
bn_eps : float, default 1e-5
Small float added to variance in Batch norm.
activation : function or str or None, default nn.ReLU(inplace=True)
Activation function or name of activation function.
"""
return ConvBlock(in_channels=in_channels, out_channels=out_channels,
kernel_size=3, stride=stride, padding=padding, dilation=dilation,
groups=groups, bias=bias, use_bn=use_bn, bn_eps=bn_eps, activation=
activation)
class HSigmoid(nn.Module):
"""
Approximated sigmoid function, so-called hard-version of sigmoid from 'Searching for MobileNetV3,'
https://arxiv.org/abs/1905.02244.
"""
def forward(self, x):
return F.relu6(x + 3.0, inplace=True) / 6.0
class HSwish(nn.Module):
"""
H-Swish activation function from 'Searching for MobileNetV3,' https://arxiv.org/abs/1905.02244.
Parameters:
----------
inplace : bool
Whether to use inplace version of the module.
"""
def __init__(self, inplace=False):
super(HSwish, self).__init__()
self.inplace = inplace
def forward(self, x):
return x * F.relu6(x + 3.0, inplace=self.inplace) / 6.0
class Swish(nn.Module):
"""
Swish activation function from 'Searching for Activation Functions,' https://arxiv.org/abs/1710.05941.
"""
def forward(self, x):
return x * torch.sigmoid(x)
class ConvBlock(nn.Module):
"""
Standard convolution block with Batch normalization and activation.
Parameters:
----------
in_channels : int
Number of input channels.
out_channels : int
Number of output channels.
kernel_size : int or tuple/list of 2 int
Convolution window size.
stride : int or tuple/list of 2 int
Strides of the convolution.
padding : int or tuple/list of 2 int
Padding value for convolution layer.
dilation : int or tuple/list of 2 int, default 1
Dilation value for convolution layer.
groups : int, default 1
Number of groups.
bias : bool, default False
Whether the layer uses a bias vector.
use_bn : bool, default True
Whether to use BatchNorm layer.
bn_eps : float, default 1e-5
Small float added to variance in Batch norm.
activation : function or str or None, default nn.ReLU(inplace=True)
Activation function or name of activation function.
"""
def __init__(self, in_channels, out_channels, kernel_size, stride,
padding, dilation=1, groups=1, bias=False, use_bn=True, bn_eps=
1e-05, activation=lambda : nn.ReLU(inplace=True)):
super(ConvBlock, self).__init__()
self.activate = activation is not None
self.use_bn = use_bn
self.conv = nn.Conv2d(in_channels=in_channels, out_channels=
out_channels, kernel_size=kernel_size, stride=stride, padding=
padding, dilation=dilation, groups=groups, bias=bias)
if self.use_bn:
self.bn = nn.BatchNorm2d(num_features=out_channels, eps=bn_eps)
if self.activate:
self.activ = get_activation_layer(activation)
def forward(self, x):
x = self.conv(x)
if self.use_bn:
x = self.bn(x)
if self.activate:
x = self.activ(x)
return x
class SPHead(nn.Module):
"""
SuperPointNet head block.
Parameters:
----------
in_channels : int
Number of input channels.
mid_channels : int
Number of middle channels.
out_channels : int
Number of output channels.
"""
def __init__(self, in_channels, mid_channels, out_channels):
super(SPHead, self).__init__()
self.conv1 = conv3x3_block(in_channels=in_channels, out_channels=
mid_channels, bias=True, use_bn=False)
self.conv2 = conv1x1(in_channels=mid_channels, out_channels=
out_channels, bias=True)
def forward(self, x):
x = self.conv1(x)
x = self.conv2(x)
return x
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'in_channels': 4, 'mid_channels': 4, 'out_channels': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
import torch.nn as nn
import torch.nn.functional as F
from inspect import isfunction
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
@triton.jit
def triton_poi_fused_convolution_relu_0(in_out_ptr0, in_ptr0, xnumel,
XBLOCK: tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = xindex // 16 % 4
tmp0 = tl.load(in_out_ptr0 + x3, xmask)
tmp1 = tl.load(in_ptr0 + x1, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tl.store(in_out_ptr0 + x3, tmp4, xmask)
@triton.jit
def triton_poi_fused_convolution_1(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl
.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = xindex // 16 % 4
tmp0 = tl.load(in_out_ptr0 + x3, xmask)
tmp1 = tl.load(in_ptr0 + x1, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + x3, tmp2, xmask)
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 3, 3), (36, 9, 3, 1))
assert_size_stride(primals_2, (4,), (1,))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_4, (4, 4, 1, 1), (4, 1, 1, 1))
assert_size_stride(primals_5, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = extern_kernels.convolution(primals_3, primals_1, stride=(1,
1), padding=(1, 1), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf0, (4, 4, 4, 4), (64, 16, 4, 1))
buf1 = buf0
del buf0
get_raw_stream(0)
triton_poi_fused_convolution_relu_0[grid(256)](buf1, primals_2, 256,
XBLOCK=128, num_warps=4, num_stages=1)
del primals_2
buf2 = extern_kernels.convolution(buf1, primals_4, stride=(1, 1),
padding=(0, 0), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf2, (4, 4, 4, 4), (64, 16, 4, 1))
buf3 = buf2
del buf2
triton_poi_fused_convolution_1[grid(256)](buf3, primals_5, 256,
XBLOCK=256, num_warps=4, num_stages=1)
del primals_5
return buf3, primals_1, primals_3, primals_4, buf1
def conv1x1(in_channels, out_channels, stride=1, groups=1, bias=False):
"""
Convolution 1x1 layer.
Parameters:
----------
in_channels : int
Number of input channels.
out_channels : int
Number of output channels.
stride : int or tuple/list of 2 int, default 1
Strides of the convolution.
groups : int, default 1
Number of groups.
bias : bool, default False
Whether the layer uses a bias vector.
"""
return nn.Conv2d(in_channels=in_channels, out_channels=out_channels,
kernel_size=1, stride=stride, groups=groups, bias=bias)
def get_activation_layer(activation):
"""
Create activation layer from string/function.
Parameters:
----------
activation : function, or str, or nn.Module
Activation function or name of activation function.
Returns
-------
nn.Module
Activation layer.
"""
assert activation is not None
if isfunction(activation):
return activation()
elif isinstance(activation, str):
if activation == 'relu':
return nn.ReLU(inplace=True)
elif activation == 'relu6':
return nn.ReLU6(inplace=True)
elif activation == 'swish':
return Swish()
elif activation == 'hswish':
return HSwish(inplace=True)
elif activation == 'sigmoid':
return nn.Sigmoid()
elif activation == 'hsigmoid':
return HSigmoid()
else:
raise NotImplementedError()
else:
assert isinstance(activation, nn.Module)
return activation
def conv3x3_block(in_channels, out_channels, stride=1, padding=1, dilation=
1, groups=1, bias=False, use_bn=True, bn_eps=1e-05, activation=lambda :
nn.ReLU(inplace=True)):
"""
3x3 version of the standard convolution block.
Parameters:
----------
in_channels : int
Number of input channels.
out_channels : int
Number of output channels.
stride : int or tuple/list of 2 int, default 1
Strides of the convolution.
padding : int or tuple/list of 2 int, default 1
Padding value for convolution layer.
dilation : int or tuple/list of 2 int, default 1
Dilation value for convolution layer.
groups : int, default 1
Number of groups.
bias : bool, default False
Whether the layer uses a bias vector.
use_bn : bool, default True
Whether to use BatchNorm layer.
bn_eps : float, default 1e-5
Small float added to variance in Batch norm.
activation : function or str or None, default nn.ReLU(inplace=True)
Activation function or name of activation function.
"""
return ConvBlock(in_channels=in_channels, out_channels=out_channels,
kernel_size=3, stride=stride, padding=padding, dilation=dilation,
groups=groups, bias=bias, use_bn=use_bn, bn_eps=bn_eps, activation=
activation)
class HSigmoid(nn.Module):
"""
Approximated sigmoid function, so-called hard-version of sigmoid from 'Searching for MobileNetV3,'
https://arxiv.org/abs/1905.02244.
"""
def forward(self, x):
return F.relu6(x + 3.0, inplace=True) / 6.0
class HSwish(nn.Module):
"""
H-Swish activation function from 'Searching for MobileNetV3,' https://arxiv.org/abs/1905.02244.
Parameters:
----------
inplace : bool
Whether to use inplace version of the module.
"""
def __init__(self, inplace=False):
super(HSwish, self).__init__()
self.inplace = inplace
def forward(self, x):
return x * F.relu6(x + 3.0, inplace=self.inplace) / 6.0
class Swish(nn.Module):
"""
Swish activation function from 'Searching for Activation Functions,' https://arxiv.org/abs/1710.05941.
"""
def forward(self, x):
return x * torch.sigmoid(x)
class ConvBlock(nn.Module):
"""
Standard convolution block with Batch normalization and activation.
Parameters:
----------
in_channels : int
Number of input channels.
out_channels : int
Number of output channels.
kernel_size : int or tuple/list of 2 int
Convolution window size.
stride : int or tuple/list of 2 int
Strides of the convolution.
padding : int or tuple/list of 2 int
Padding value for convolution layer.
dilation : int or tuple/list of 2 int, default 1
Dilation value for convolution layer.
groups : int, default 1
Number of groups.
bias : bool, default False
Whether the layer uses a bias vector.
use_bn : bool, default True
Whether to use BatchNorm layer.
bn_eps : float, default 1e-5
Small float added to variance in Batch norm.
activation : function or str or None, default nn.ReLU(inplace=True)
Activation function or name of activation function.
"""
def __init__(self, in_channels, out_channels, kernel_size, stride,
padding, dilation=1, groups=1, bias=False, use_bn=True, bn_eps=
1e-05, activation=lambda : nn.ReLU(inplace=True)):
super(ConvBlock, self).__init__()
self.activate = activation is not None
self.use_bn = use_bn
self.conv = nn.Conv2d(in_channels=in_channels, out_channels=
out_channels, kernel_size=kernel_size, stride=stride, padding=
padding, dilation=dilation, groups=groups, bias=bias)
if self.use_bn:
self.bn = nn.BatchNorm2d(num_features=out_channels, eps=bn_eps)
if self.activate:
self.activ = get_activation_layer(activation)
def forward(self, x):
x = self.conv(x)
if self.use_bn:
x = self.bn(x)
if self.activate:
x = self.activ(x)
return x
class SPHeadNew(nn.Module):
"""
SuperPointNet head block.
Parameters:
----------
in_channels : int
Number of input channels.
mid_channels : int
Number of middle channels.
out_channels : int
Number of output channels.
"""
def __init__(self, in_channels, mid_channels, out_channels):
super(SPHeadNew, self).__init__()
self.conv1 = conv3x3_block(in_channels=in_channels, out_channels=
mid_channels, bias=True, use_bn=False)
self.conv2 = conv1x1(in_channels=mid_channels, out_channels=
out_channels, bias=True)
def forward(self, input_0):
primals_1 = self.conv1.conv.weight
primals_2 = self.conv1.conv.bias
primals_4 = self.conv2.weight
primals_5 = self.conv2.bias
primals_3 = input_0
output = call([primals_1, primals_2, primals_3, primals_4, primals_5])
return output[0]
| iofthetiger/pkuad | SPHead | false | 6,902 | [
"Apache-2.0"
] | 1 | 07496d108c614c84be028f344830becc9cac8fe5 | https://github.com/iofthetiger/pkuad/tree/07496d108c614c84be028f344830becc9cac8fe5 | import torch
import torch.nn as nn
import torch.nn.functional as F
from inspect import isfunction
def conv1x1(in_channels, out_channels, stride=1, groups=1, bias=False):
"""
Convolution 1x1 layer.
Parameters:
----------
in_channels : int
Number of input channels.
out_channels : int
Number of output channels.
stride : int or tuple/list of 2 int, default 1
Strides of the convolution.
groups : int, default 1
Number of groups.
bias : bool, default False
Whether the layer uses a bias vector.
"""
return nn.Conv2d(in_channels=in_channels, out_channels=out_channels,
kernel_size=1, stride=stride, groups=groups, bias=bias)
def get_activation_layer(activation):
"""
Create activation layer from string/function.
Parameters:
----------
activation : function, or str, or nn.Module
Activation function or name of activation function.
Returns
-------
nn.Module
Activation layer.
"""
assert activation is not None
if isfunction(activation):
return activation()
elif isinstance(activation, str):
if activation == 'relu':
return nn.ReLU(inplace=True)
elif activation == 'relu6':
return nn.ReLU6(inplace=True)
elif activation == 'swish':
return Swish()
elif activation == 'hswish':
return HSwish(inplace=True)
elif activation == 'sigmoid':
return nn.Sigmoid()
elif activation == 'hsigmoid':
return HSigmoid()
else:
raise NotImplementedError()
else:
assert isinstance(activation, nn.Module)
return activation
def conv3x3_block(in_channels, out_channels, stride=1, padding=1, dilation=
1, groups=1, bias=False, use_bn=True, bn_eps=1e-05, activation=lambda :
nn.ReLU(inplace=True)):
"""
3x3 version of the standard convolution block.
Parameters:
----------
in_channels : int
Number of input channels.
out_channels : int
Number of output channels.
stride : int or tuple/list of 2 int, default 1
Strides of the convolution.
padding : int or tuple/list of 2 int, default 1
Padding value for convolution layer.
dilation : int or tuple/list of 2 int, default 1
Dilation value for convolution layer.
groups : int, default 1
Number of groups.
bias : bool, default False
Whether the layer uses a bias vector.
use_bn : bool, default True
Whether to use BatchNorm layer.
bn_eps : float, default 1e-5
Small float added to variance in Batch norm.
activation : function or str or None, default nn.ReLU(inplace=True)
Activation function or name of activation function.
"""
return ConvBlock(in_channels=in_channels, out_channels=out_channels,
kernel_size=3, stride=stride, padding=padding, dilation=dilation,
groups=groups, bias=bias, use_bn=use_bn, bn_eps=bn_eps, activation=
activation)
class HSigmoid(nn.Module):
"""
Approximated sigmoid function, so-called hard-version of sigmoid from 'Searching for MobileNetV3,'
https://arxiv.org/abs/1905.02244.
"""
def forward(self, x):
return F.relu6(x + 3.0, inplace=True) / 6.0
class HSwish(nn.Module):
"""
H-Swish activation function from 'Searching for MobileNetV3,' https://arxiv.org/abs/1905.02244.
Parameters:
----------
inplace : bool
Whether to use inplace version of the module.
"""
def __init__(self, inplace=False):
super().__init__()
self.inplace = inplace
def forward(self, x):
return x * F.relu6(x + 3.0, inplace=self.inplace) / 6.0
class Swish(nn.Module):
"""
Swish activation function from 'Searching for Activation Functions,' https://arxiv.org/abs/1710.05941.
"""
def forward(self, x):
return x * torch.sigmoid(x)
class ConvBloc
# ... truncated (>4000 chars) for memory efficiency |
NTXent | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/zk/czk5xfokmwnuegxn53eciq25366p2is3a6lxx47tlosf3q225vha.py
# Topologically Sorted Source Nodes: [z1], Original ATen: [aten.div]
# Source node to ATen node mapping:
# z1 => div
# Graph fragment:
# %div : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%arg0_1, %expand), kwargs = {})
triton_poi_fused_div_0 = async_compile.triton('triton_poi_fused_div_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_div_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_div_0(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp9 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp2 = tmp1 * tmp1
tmp4 = tmp3 * tmp3
tmp5 = tmp2 + tmp4
tmp7 = tmp6 * tmp6
tmp8 = tmp5 + tmp7
tmp10 = tmp9 * tmp9
tmp11 = tmp8 + tmp10
tmp12 = libdevice.sqrt(tmp11)
tmp13 = 1e-12
tmp14 = triton_helpers.maximum(tmp12, tmp13)
tmp15 = tmp0 / tmp14
tl.store(out_ptr0 + (x2), tmp15, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/lg/clgz3btdtfw2lht5p3llry2elisvzugzjxrsi555c5gk3rbomt63.py
# Topologically Sorted Source Nodes: [targets, loss, exp, similarity_1], Original ATen: [aten.arange, aten.nll_loss_forward, aten.exp, aten.mul, aten._log_softmax]
# Source node to ATen node mapping:
# exp => exp
# loss => amax, convert_element_type, div_2, exp_1, full_default_1, ne_1, ne_2, neg, sub, sum_3, sum_4, sum_5, where_1
# similarity_1 => mul
# targets => iota
# Graph fragment:
# %iota : [num_users=4] = call_function[target=torch.ops.prims.iota.default](args = (4,), kwargs = {start: 0, step: 1, dtype: torch.int64, device: cuda:0, requires_grad: False})
# %ne_1 : [num_users=1] = call_function[target=torch.ops.aten.ne.Scalar](args = (%iota, -100), kwargs = {})
# %exp : [num_users=1] = call_function[target=torch.ops.aten.exp.default](args = (%arg2_1,), kwargs = {})
# %mul : [num_users=2] = call_function[target=torch.ops.aten.mul.Tensor](args = (%mm, %exp), kwargs = {})
# %amax : [num_users=1] = call_function[target=torch.ops.aten.amax.default](args = (%mul, [1], True), kwargs = {})
# %sub : [num_users=2] = call_function[target=torch.ops.aten.sub.Tensor](args = (%mul, %amax), kwargs = {})
# %exp_1 : [num_users=1] = call_function[target=torch.ops.aten.exp.default](args = (%sub,), kwargs = {})
# %sum_3 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%exp_1, [1], True), kwargs = {})
# %neg : [num_users=1] = call_function[target=torch.ops.aten.neg.default](args = (%squeeze,), kwargs = {})
# %full_default_1 : [num_users=1] = call_function[target=torch.ops.aten.full.default](args = ([], 0.0), kwargs = {dtype: torch.float32, layout: torch.strided, device: cuda:0, pin_memory: False})
# %where_1 : [num_users=1] = call_function[target=torch.ops.aten.where.self](args = (%ne_1, %neg, %full_default_1), kwargs = {})
# %sum_5 : [num_users=1] = call_function[target=torch.ops.aten.sum.default](args = (%where_1,), kwargs = {})
# %ne_2 : [num_users=1] = call_function[target=torch.ops.aten.ne.Scalar](args = (%iota, -100), kwargs = {})
# %sum_4 : [num_users=1] = call_function[target=torch.ops.aten.sum.default](args = (%ne_2,), kwargs = {})
# %convert_element_type : [num_users=1] = call_function[target=torch.ops.prims.convert_element_type.default](args = (%sum_4, torch.float32), kwargs = {})
# %div_2 : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%sum_5, %convert_element_type), kwargs = {})
triton_per_fused__log_softmax_arange_exp_mul_nll_loss_forward_1 = async_compile.triton('triton_per_fused__log_softmax_arange_exp_mul_nll_loss_forward_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[1, 4],
reduction_hint=ReductionHint.DEFAULT,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {3: 1}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=(3,))]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused__log_softmax_arange_exp_mul_nll_loss_forward_1', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 8, 'num_reduction': 2, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused__log_softmax_arange_exp_mul_nll_loss_forward_1(in_out_ptr0, in_ptr0, in_ptr1, xnumel, rnumel, XBLOCK : tl.constexpr):
xnumel = 1
rnumel = 4
RBLOCK: tl.constexpr = 4
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[None, :]
roffset = 0
rmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
r0 = rindex
tmp0 = tl.load(in_ptr0 + (4*r0), None, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (4*r0), None, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (1 + (4*r0)), None, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr1 + (1 + (4*r0)), None, eviction_policy='evict_last')
tmp9 = tl.load(in_ptr0 + (2 + (4*r0)), None, eviction_policy='evict_last')
tmp10 = tl.load(in_ptr1 + (2 + (4*r0)), None, eviction_policy='evict_last')
tmp14 = tl.load(in_ptr0 + (3 + (4*r0)), None, eviction_policy='evict_last')
tmp15 = tl.load(in_ptr1 + (3 + (4*r0)), None, eviction_policy='evict_last')
tmp2 = tl_math.exp(tmp1)
tmp3 = tmp0 * tmp2
tmp6 = tl_math.exp(tmp5)
tmp7 = tmp4 * tmp6
tmp8 = triton_helpers.maximum(tmp3, tmp7)
tmp11 = tl_math.exp(tmp10)
tmp12 = tmp9 * tmp11
tmp13 = triton_helpers.maximum(tmp8, tmp12)
tmp16 = tl_math.exp(tmp15)
tmp17 = tmp14 * tmp16
tmp18 = triton_helpers.maximum(tmp13, tmp17)
tmp19 = tmp3 - tmp18
tmp20 = tl_math.exp(tmp19)
tmp21 = tmp7 - tmp18
tmp22 = tl_math.exp(tmp21)
tmp23 = tmp20 + tmp22
tmp24 = tmp12 - tmp18
tmp25 = tl_math.exp(tmp24)
tmp26 = tmp23 + tmp25
tmp27 = tmp17 - tmp18
tmp28 = tl_math.exp(tmp27)
tmp29 = tmp26 + tmp28
tmp30 = r0
tmp31 = tl.full([1, 1], -100, tl.int64)
tmp32 = tmp30 != tmp31
tmp33 = tl.full([1, 1], 0, tl.int64)
tmp34 = tl.where(tmp32, tmp30, tmp33)
tmp35 = tl.load(in_ptr0 + (tmp34 + (4*r0)), None, eviction_policy='evict_last')
tmp36 = tl.load(in_ptr1 + (tmp34 + (4*r0)), None, eviction_policy='evict_last')
tmp37 = tl_math.exp(tmp36)
tmp38 = tmp35 * tmp37
tmp39 = tmp38 - tmp18
tmp40 = tl_math.log(tmp29)
tmp41 = tmp39 - tmp40
tmp42 = -tmp41
tmp43 = 0.0
tmp44 = tl.where(tmp32, tmp42, tmp43)
tmp45 = tl.broadcast_to(tmp44, [XBLOCK, RBLOCK])
tmp47 = tl.sum(tmp45, 1)[:, None]
tmp48 = tmp32.to(tl.int64)
tmp49 = tl.broadcast_to(tmp48, [XBLOCK, RBLOCK])
tmp51 = tl.sum(tmp49, 1)[:, None]
tmp52 = tmp51.to(tl.float32)
tmp53 = tmp47 / tmp52
tl.debug_barrier()
tl.store(in_out_ptr0 + (tl.full([XBLOCK, 1], 0, tl.int32)), tmp53, None)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, arg1_1, arg2_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4), (4, 1))
assert_size_stride(arg1_1, (4, 4), (4, 1))
assert_size_stride(arg2_1, (4, 4), (4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [z1], Original ATen: [aten.div]
stream0 = get_raw_stream(0)
triton_poi_fused_div_0.run(arg0_1, buf0, 16, grid=grid(16), stream=stream0)
del arg0_1
buf1 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [z2], Original ATen: [aten.div]
triton_poi_fused_div_0.run(arg1_1, buf1, 16, grid=grid(16), stream=stream0)
del arg1_1
buf2 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [z1, similarity], Original ATen: [aten.div, aten.mm]
extern_kernels.mm(buf0, reinterpret_tensor(buf1, (4, 4), (1, 4), 0), out=buf2)
del buf0
del buf1
buf5 = empty_strided_cuda((), (), torch.float32)
buf7 = buf5; del buf5 # reuse
# Topologically Sorted Source Nodes: [targets, loss, exp, similarity_1], Original ATen: [aten.arange, aten.nll_loss_forward, aten.exp, aten.mul, aten._log_softmax]
triton_per_fused__log_softmax_arange_exp_mul_nll_loss_forward_1.run(buf7, buf2, arg2_1, 1, 4, grid=grid(1), stream=stream0)
del arg2_1
del buf2
return (buf7, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
arg1_1 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
arg2_1 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([arg0_1, arg1_1, arg2_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.nn.functional as F
class NTXent(nn.Module):
def forward(self, z1, z2, t):
batch_size = z1.shape[0]
device = z1.device
z1 = F.normalize(z1, dim=-1)
z2 = F.normalize(z2, dim=-1)
similarity = torch.matmul(z1, z2.T)
similarity = similarity * torch.exp(t)
targets = torch.arange(batch_size, device=device)
loss = F.cross_entropy(similarity, targets)
return loss
def get_inputs():
return [torch.rand([4, 4]), torch.rand([4, 4]), torch.rand([4, 4])]
def get_init_inputs():
return [[], {}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_div_0(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp9 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp2 = tmp1 * tmp1
tmp4 = tmp3 * tmp3
tmp5 = tmp2 + tmp4
tmp7 = tmp6 * tmp6
tmp8 = tmp5 + tmp7
tmp10 = tmp9 * tmp9
tmp11 = tmp8 + tmp10
tmp12 = libdevice.sqrt(tmp11)
tmp13 = 1e-12
tmp14 = triton_helpers.maximum(tmp12, tmp13)
tmp15 = tmp0 / tmp14
tl.store(out_ptr0 + x2, tmp15, xmask)
@triton.jit
def triton_per_fused__log_softmax_arange_exp_mul_nll_loss_forward_1(in_out_ptr0
, in_ptr0, in_ptr1, xnumel, rnumel, XBLOCK: tl.constexpr):
RBLOCK: tl.constexpr = 4
xoffset = tl.program_id(0) * XBLOCK
xoffset + tl.arange(0, XBLOCK)[:, None]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[None, :]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
r0 = rindex
tmp0 = tl.load(in_ptr0 + 4 * r0, None, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + 4 * r0, None, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (1 + 4 * r0), None, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr1 + (1 + 4 * r0), None, eviction_policy='evict_last')
tmp9 = tl.load(in_ptr0 + (2 + 4 * r0), None, eviction_policy='evict_last')
tmp10 = tl.load(in_ptr1 + (2 + 4 * r0), None, eviction_policy='evict_last')
tmp14 = tl.load(in_ptr0 + (3 + 4 * r0), None, eviction_policy='evict_last')
tmp15 = tl.load(in_ptr1 + (3 + 4 * r0), None, eviction_policy='evict_last')
tmp2 = tl_math.exp(tmp1)
tmp3 = tmp0 * tmp2
tmp6 = tl_math.exp(tmp5)
tmp7 = tmp4 * tmp6
tmp8 = triton_helpers.maximum(tmp3, tmp7)
tmp11 = tl_math.exp(tmp10)
tmp12 = tmp9 * tmp11
tmp13 = triton_helpers.maximum(tmp8, tmp12)
tmp16 = tl_math.exp(tmp15)
tmp17 = tmp14 * tmp16
tmp18 = triton_helpers.maximum(tmp13, tmp17)
tmp19 = tmp3 - tmp18
tmp20 = tl_math.exp(tmp19)
tmp21 = tmp7 - tmp18
tmp22 = tl_math.exp(tmp21)
tmp23 = tmp20 + tmp22
tmp24 = tmp12 - tmp18
tmp25 = tl_math.exp(tmp24)
tmp26 = tmp23 + tmp25
tmp27 = tmp17 - tmp18
tmp28 = tl_math.exp(tmp27)
tmp29 = tmp26 + tmp28
tmp30 = r0
tmp31 = tl.full([1, 1], -100, tl.int64)
tmp32 = tmp30 != tmp31
tmp33 = tl.full([1, 1], 0, tl.int64)
tmp34 = tl.where(tmp32, tmp30, tmp33)
tmp35 = tl.load(in_ptr0 + (tmp34 + 4 * r0), None, eviction_policy=
'evict_last')
tmp36 = tl.load(in_ptr1 + (tmp34 + 4 * r0), None, eviction_policy=
'evict_last')
tmp37 = tl_math.exp(tmp36)
tmp38 = tmp35 * tmp37
tmp39 = tmp38 - tmp18
tmp40 = tl_math.log(tmp29)
tmp41 = tmp39 - tmp40
tmp42 = -tmp41
tmp43 = 0.0
tmp44 = tl.where(tmp32, tmp42, tmp43)
tmp45 = tl.broadcast_to(tmp44, [XBLOCK, RBLOCK])
tmp47 = tl.sum(tmp45, 1)[:, None]
tmp48 = tmp32.to(tl.int64)
tmp49 = tl.broadcast_to(tmp48, [XBLOCK, RBLOCK])
tmp51 = tl.sum(tmp49, 1)[:, None]
tmp52 = tmp51.to(tl.float32)
tmp53 = tmp47 / tmp52
tl.debug_barrier()
tl.store(in_out_ptr0 + tl.full([XBLOCK, 1], 0, tl.int32), tmp53, None)
def call(args):
arg0_1, arg1_1, arg2_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4), (4, 1))
assert_size_stride(arg1_1, (4, 4), (4, 1))
assert_size_stride(arg2_1, (4, 4), (4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_div_0[grid(16)](arg0_1, buf0, 16, XBLOCK=16,
num_warps=1, num_stages=1)
del arg0_1
buf1 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
triton_poi_fused_div_0[grid(16)](arg1_1, buf1, 16, XBLOCK=16,
num_warps=1, num_stages=1)
del arg1_1
buf2 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.mm(buf0, reinterpret_tensor(buf1, (4, 4), (1, 4), 0),
out=buf2)
del buf0
del buf1
buf5 = empty_strided_cuda((), (), torch.float32)
buf7 = buf5
del buf5
triton_per_fused__log_softmax_arange_exp_mul_nll_loss_forward_1[grid(1)
](buf7, buf2, arg2_1, 1, 4, XBLOCK=1, num_warps=2, num_stages=1)
del arg2_1
del buf2
return buf7,
class NTXentNew(nn.Module):
def forward(self, input_0, input_1, input_2):
arg0_1 = input_0
arg1_1 = input_1
arg2_1 = input_2
output = call([arg0_1, arg1_1, arg2_1])
return output[0]
| isaaccorley/contrastive-surface-image-pretraining | NTXent | false | 6,903 | [
"MIT"
] | 1 | a918d4fd3b9cc61ec512af978fb4f086d3b46a70 | https://github.com/isaaccorley/contrastive-surface-image-pretraining/tree/a918d4fd3b9cc61ec512af978fb4f086d3b46a70 | import torch
import torch.nn as nn
import torch.nn.functional as F
class Model(nn.Module):
def forward(self, z1, z2, t):
batch_size = z1.shape[0]
device = z1.device
z1 = F.normalize(z1, dim=-1)
z2 = F.normalize(z2, dim=-1)
similarity = torch.matmul(z1, z2.T)
similarity = similarity * torch.exp(t)
targets = torch.arange(batch_size, device=device)
loss = F.cross_entropy(similarity, targets)
return loss
def get_inputs():
return [torch.rand([4, 4]), torch.rand([4, 4]), torch.rand([4, 4])]
def get_init_inputs():
return []
|
VectorQuantizer | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/x6/cx6vj6jd2qhcbwnwwajpsdipepno6iepsyrvmepfo457t522sl26.py
# Topologically Sorted Source Nodes: [inputs, flat_input], Original ATen: [aten.clone, aten.view]
# Source node to ATen node mapping:
# flat_input => view
# inputs => clone
# Graph fragment:
# %clone : [num_users=3] = call_function[target=torch.ops.aten.clone.default](args = (%permute,), kwargs = {memory_format: torch.contiguous_format})
# %view : [num_users=2] = call_function[target=torch.ops.aten.reshape.default](args = (%clone, [-1, 4]), kwargs = {})
triton_poi_fused_clone_view_0 = async_compile.triton('triton_poi_fused_clone_view_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64, 4], tile_hint=TileHint.SQUARE,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clone_view_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clone_view_0(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 64
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x1 = xindex
y0 = yindex
tmp0 = tl.load(in_ptr0 + ((16*x1) + (64*(y0 // 16)) + (y0 % 16)), xmask & ymask, eviction_policy='evict_last')
tl.store(out_ptr0 + (x1 + (4*y0)), tmp0, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/oh/coh3kyu6rghphyqtfflzqlsbjfinc4c7cmryrm7sbhmrnnvqqu6w.py
# Topologically Sorted Source Nodes: [pow_1, sum_1, pow_2, sum_2, add, mul, distances], Original ATen: [aten.pow, aten.sum, aten.add, aten.mul, aten.sub]
# Source node to ATen node mapping:
# add => add
# distances => sub
# mul => mul
# pow_1 => pow_1
# pow_2 => pow_2
# sum_1 => sum_1
# sum_2 => sum_2
# Graph fragment:
# %pow_1 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%view, 2), kwargs = {})
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%pow_1, [1], True), kwargs = {})
# %pow_2 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%primals_2, 2), kwargs = {})
# %sum_2 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%pow_2, [1]), kwargs = {})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%sum_1, %sum_2), kwargs = {})
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%mm, 2), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%add, %mul), kwargs = {})
triton_poi_fused_add_mul_pow_sub_sum_1 = async_compile.triton('triton_poi_fused_add_mul_pow_sub_sum_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_mul_pow_sub_sum_1', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 9, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_mul_pow_sub_sum_1(in_out_ptr0, in_ptr0, in_ptr1, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = (xindex // 4)
x0 = xindex % 4
x2 = xindex
tmp0 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp8 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp11 = tl.load(in_ptr1 + (4*x0), xmask, eviction_policy='evict_last')
tmp13 = tl.load(in_ptr1 + (1 + (4*x0)), xmask, eviction_policy='evict_last')
tmp16 = tl.load(in_ptr1 + (2 + (4*x0)), xmask, eviction_policy='evict_last')
tmp19 = tl.load(in_ptr1 + (3 + (4*x0)), xmask, eviction_policy='evict_last')
tmp23 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tmp0 * tmp0
tmp3 = tmp2 * tmp2
tmp4 = tmp1 + tmp3
tmp6 = tmp5 * tmp5
tmp7 = tmp4 + tmp6
tmp9 = tmp8 * tmp8
tmp10 = tmp7 + tmp9
tmp12 = tmp11 * tmp11
tmp14 = tmp13 * tmp13
tmp15 = tmp12 + tmp14
tmp17 = tmp16 * tmp16
tmp18 = tmp15 + tmp17
tmp20 = tmp19 * tmp19
tmp21 = tmp18 + tmp20
tmp22 = tmp10 + tmp21
tmp24 = 2.0
tmp25 = tmp23 * tmp24
tmp26 = tmp22 - tmp25
tl.store(in_out_ptr0 + (x2), tmp26, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/3l/c3ltycq3kz4scgm57x763xs5lwirtcz6gwjv67cofz752mk5hx4y.py
# Topologically Sorted Source Nodes: [argmin], Original ATen: [aten.argmin]
# Source node to ATen node mapping:
# argmin => argmin
# Graph fragment:
# %argmin : [num_users=1] = call_function[target=torch.ops.aten.argmin.default](args = (%sub, 1), kwargs = {})
triton_poi_fused_argmin_2 = async_compile.triton('triton_poi_fused_argmin_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*i64', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_argmin_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 4, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_argmin_2(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (4*x0), xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr0 + (1 + (4*x0)), xmask, eviction_policy='evict_last')
tmp17 = tl.load(in_ptr0 + (2 + (4*x0)), xmask, eviction_policy='evict_last')
tmp32 = tl.load(in_ptr0 + (3 + (4*x0)), xmask, eviction_policy='evict_last')
tmp2 = tmp0 < tmp1
tmp3 = tmp0 == tmp1
tmp4 = tmp0 != tmp0
tmp5 = tmp1 != tmp1
tmp6 = tmp4 > tmp5
tmp7 = tmp2 | tmp6
tmp8 = tmp4 & tmp5
tmp9 = tmp3 | tmp8
tmp10 = tl.full([1], 0, tl.int64)
tmp11 = tl.full([1], 1, tl.int64)
tmp12 = tmp10 < tmp11
tmp13 = tmp9 & tmp12
tmp14 = tmp7 | tmp13
tmp15 = tl.where(tmp14, tmp0, tmp1)
tmp16 = tl.where(tmp14, tmp10, tmp11)
tmp18 = tmp15 < tmp17
tmp19 = tmp15 == tmp17
tmp20 = tmp15 != tmp15
tmp21 = tmp17 != tmp17
tmp22 = tmp20 > tmp21
tmp23 = tmp18 | tmp22
tmp24 = tmp20 & tmp21
tmp25 = tmp19 | tmp24
tmp26 = tl.full([1], 2, tl.int64)
tmp27 = tmp16 < tmp26
tmp28 = tmp25 & tmp27
tmp29 = tmp23 | tmp28
tmp30 = tl.where(tmp29, tmp15, tmp17)
tmp31 = tl.where(tmp29, tmp16, tmp26)
tmp33 = tmp30 < tmp32
tmp34 = tmp30 == tmp32
tmp35 = tmp30 != tmp30
tmp36 = tmp32 != tmp32
tmp37 = tmp35 > tmp36
tmp38 = tmp33 | tmp37
tmp39 = tmp35 & tmp36
tmp40 = tmp34 | tmp39
tmp41 = tl.full([1], 3, tl.int64)
tmp42 = tmp31 < tmp41
tmp43 = tmp40 & tmp42
tmp44 = tmp38 | tmp43
tmp45 = tl.where(tmp44, tmp30, tmp32)
tmp46 = tl.where(tmp44, tmp31, tmp41)
tl.store(out_ptr0 + (x0), tmp46, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/rl/crlo3ugzwljr2rgpi3zcuc6lfaljqn4v44qgasneddceseggjqyz.py
# Topologically Sorted Source Nodes: [scatter_], Original ATen: [aten.scatter]
# Source node to ATen node mapping:
# scatter_ => scatter_upon_const_tensor
# Graph fragment:
# %scatter_upon_const_tensor : [num_users=4] = call_function[target=torch._inductor.fx_passes.post_grad.scatter_upon_const_tensor](args = (), kwargs = {shape: [64, 4], background_val: 0, dtype: torch.float32, dim: 1, selector: %unsqueeze, val: 1})
triton_poi_fused_scatter_3 = async_compile.triton('triton_poi_fused_scatter_3', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*i64', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_scatter_3', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_scatter_3(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = (xindex // 4)
x0 = xindex % 4
x2 = xindex
tmp0 = tl.load(in_ptr0 + (x1), xmask, eviction_policy='evict_last')
tmp1 = x0
tmp2 = tmp0 == tmp1
tmp3 = 1.0
tmp4 = 0.0
tmp5 = tl.where(tmp2, tmp3, tmp4)
tl.store(out_ptr0 + (x2), tmp5, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/rl/crlh4w24igjub75ofntngq2kxti3e545jiew4yq2ywudajjeo6rg.py
# Topologically Sorted Source Nodes: [inputs, e_latent_loss, mul_1, loss], Original ATen: [aten.clone, aten.mse_loss, aten.mul, aten.add]
# Source node to ATen node mapping:
# e_latent_loss => mean, pow_3, sub_1
# inputs => clone
# loss => add_1
# mul_1 => mul_1
# Graph fragment:
# %clone : [num_users=3] = call_function[target=torch.ops.aten.clone.default](args = (%permute,), kwargs = {memory_format: torch.contiguous_format})
# %sub_1 : [num_users=3] = call_function[target=torch.ops.aten.sub.Tensor](args = (%view_2, %clone), kwargs = {})
# %pow_3 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%sub_1, 2), kwargs = {})
# %mean : [num_users=2] = call_function[target=torch.ops.aten.mean.default](args = (%pow_3,), kwargs = {})
# %mul_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%mean, 4), kwargs = {})
# %add_1 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mean, %mul_1), kwargs = {})
triton_per_fused_add_clone_mse_loss_mul_4 = async_compile.triton('triton_per_fused_add_clone_mse_loss_mul_4', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[1, 256],
reduction_hint=ReductionHint.INNER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {3: 1}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 4), equal_to_1=(3,))]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused_add_clone_mse_loss_mul_4', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': True, 'num_load': 2, 'num_reduction': 1, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused_add_clone_mse_loss_mul_4(in_out_ptr0, in_ptr0, in_ptr1, xnumel, rnumel):
xnumel = 1
XBLOCK: tl.constexpr = 1
rnumel = 256
RBLOCK: tl.constexpr = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = tl.full([1], xoffset, tl.int32)
xmask = tl.full([RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[:]
roffset = 0
rmask = tl.full([RBLOCK], True, tl.int1)
r3 = rindex
r0 = rindex % 4
r1 = (rindex // 4) % 16
r2 = (rindex // 64)
tmp0 = tl.load(in_ptr0 + (r3), None)
tmp1 = tl.load(in_ptr1 + (r1 + (16*r0) + (64*r2)), None, eviction_policy='evict_last')
tmp2 = tmp0 - tmp1
tmp3 = tmp2 * tmp2
tmp4 = tl.broadcast_to(tmp3, [RBLOCK])
tmp6 = triton_helpers.promote_to_tensor(tl.sum(tmp4, 0))
tmp7 = 256.0
tmp8 = tmp6 / tmp7
tmp9 = 4.0
tmp10 = tmp8 * tmp9
tmp11 = tmp8 + tmp10
tl.debug_barrier()
tl.store(in_out_ptr0 + (tl.full([1], 0, tl.int32)), tmp11, None)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/a2/ca232fin2jffpjv5w2zyhuh2ckhenxab6tdn2yi57j6ei5g4qs4m.py
# Topologically Sorted Source Nodes: [avg_probs], Original ATen: [aten.mean]
# Source node to ATen node mapping:
# avg_probs => mean_2
# Graph fragment:
# %mean_2 : [num_users=2] = call_function[target=torch.ops.aten.mean.dim](args = (%scatter_upon_const_tensor, [0]), kwargs = {})
triton_per_fused_mean_5 = async_compile.triton('triton_per_fused_mean_5', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[4, 64],
reduction_hint=ReductionHint.OUTER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused_mean_5', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 1, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused_mean_5(in_ptr0, out_ptr0, xnumel, rnumel, XBLOCK : tl.constexpr):
xnumel = 4
rnumel = 64
RBLOCK: tl.constexpr = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
roffset = 0
rmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
r1 = rindex
x0 = xindex
tmp0 = tl.load(in_ptr0 + (x0 + (4*r1)), xmask, other=0.0)
tmp1 = tl.broadcast_to(tmp0, [XBLOCK, RBLOCK])
tmp3 = tl.where(xmask, tmp1, 0)
tmp4 = tl.sum(tmp3, 1)[:, None]
tl.store(out_ptr0 + (x0), tmp4, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/4w/c4wxalb74hnqm7bofzo3lg3agija2prtunuv57hmktdvopbkhsxc.py
# Topologically Sorted Source Nodes: [avg_probs, add_3, log, mul_2, sum_3, neg, perplexity], Original ATen: [aten.mean, aten.add, aten.log, aten.mul, aten.sum, aten.neg, aten.exp]
# Source node to ATen node mapping:
# add_3 => add_3
# avg_probs => mean_2
# log => log
# mul_2 => mul_2
# neg => neg
# perplexity => exp
# sum_3 => sum_3
# Graph fragment:
# %mean_2 : [num_users=2] = call_function[target=torch.ops.aten.mean.dim](args = (%scatter_upon_const_tensor, [0]), kwargs = {})
# %add_3 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mean_2, 1e-10), kwargs = {})
# %log : [num_users=1] = call_function[target=torch.ops.aten.log.default](args = (%add_3,), kwargs = {})
# %mul_2 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%mean_2, %log), kwargs = {})
# %sum_3 : [num_users=1] = call_function[target=torch.ops.aten.sum.default](args = (%mul_2,), kwargs = {})
# %neg : [num_users=1] = call_function[target=torch.ops.aten.neg.default](args = (%sum_3,), kwargs = {})
# %exp : [num_users=1] = call_function[target=torch.ops.aten.exp.default](args = (%neg,), kwargs = {})
triton_per_fused_add_exp_log_mean_mul_neg_sum_6 = async_compile.triton('triton_per_fused_add_exp_log_mean_mul_neg_sum_6', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[1, 4],
reduction_hint=ReductionHint.INNER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {2: 1}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1), equal_to_1=(2,))]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused_add_exp_log_mean_mul_neg_sum_6', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 1, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused_add_exp_log_mean_mul_neg_sum_6(in_out_ptr0, in_ptr0, xnumel, rnumel, XBLOCK : tl.constexpr):
xnumel = 1
rnumel = 4
RBLOCK: tl.constexpr = 4
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[None, :]
roffset = 0
rmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
r0 = rindex
tmp0 = tl.load(in_ptr0 + (r0), None)
tmp1 = 64.0
tmp2 = tmp0 / tmp1
tmp3 = 1e-10
tmp4 = tmp2 + tmp3
tmp5 = tl_math.log(tmp4)
tmp6 = tmp2 * tmp5
tmp7 = tl.broadcast_to(tmp6, [XBLOCK, RBLOCK])
tmp9 = tl.sum(tmp7, 1)[:, None]
tmp10 = -tmp9
tmp11 = tl_math.exp(tmp10)
tl.debug_barrier()
tl.store(in_out_ptr0 + (tl.full([XBLOCK, 1], 0, tl.int32)), tmp11, None)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/ut/cutu3atbdahfmtyqne2geeq5zl4j2hwswrriy5kdbgboi7c3aidq.py
# Topologically Sorted Source Nodes: [contiguous_1], Original ATen: [aten.clone]
# Source node to ATen node mapping:
# contiguous_1 => clone_1
# Graph fragment:
# %clone_1 : [num_users=1] = call_function[target=torch.ops.aten.clone.default](args = (%permute_2,), kwargs = {memory_format: torch.contiguous_format})
triton_poi_fused_clone_7 = async_compile.triton('triton_poi_fused_clone_7', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16, 16], tile_hint=TileHint.DEFAULT,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clone_7', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clone_7(in_ptr0, in_ptr1, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 16
xnumel = 16
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y3 = yindex
y0 = yindex % 4
y1 = (yindex // 4)
tmp0 = tl.load(in_ptr0 + (x2 + (16*y3)), xmask & ymask)
tmp1 = tl.load(in_ptr1 + (y0 + (4*x2) + (64*y1)), xmask & ymask)
tmp2 = tmp1 - tmp0
tmp3 = tmp0 + tmp2
tl.store(out_ptr0 + (x2 + (16*y3)), tmp3, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/kt/ckteflldoepzirjw6jyk3eate2sra2igoaeiohc65qghsunimlko.py
# Topologically Sorted Source Nodes: [inputs, e_latent_loss], Original ATen: [aten.clone, aten.mse_loss, aten.mse_loss_backward]
# Source node to ATen node mapping:
# e_latent_loss => sub_1
# inputs => clone
# Graph fragment:
# %clone : [num_users=3] = call_function[target=torch.ops.aten.clone.default](args = (%permute,), kwargs = {memory_format: torch.contiguous_format})
# %sub_1 : [num_users=3] = call_function[target=torch.ops.aten.sub.Tensor](args = (%view_2, %clone), kwargs = {})
# %mul_3 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sub_1, 0.0078125), kwargs = {})
triton_poi_fused_clone_mse_loss_mse_loss_backward_8 = async_compile.triton('triton_poi_fused_clone_mse_loss_mse_loss_backward_8', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64, 4], tile_hint=TileHint.SQUARE,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clone_mse_loss_mse_loss_backward_8', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clone_mse_loss_mse_loss_backward_8(in_out_ptr0, in_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 64
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y3 = yindex
y0 = yindex % 16
y1 = (yindex // 16)
tmp0 = tl.load(in_out_ptr0 + (x2 + (4*y3)), xmask & ymask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr0 + (y0 + (16*x2) + (64*y1)), xmask & ymask, eviction_policy='evict_last')
tmp2 = tmp0 - tmp1
tmp3 = 0.0078125
tmp4 = tmp2 * tmp3
tl.debug_barrier()
tl.store(in_out_ptr0 + (x2 + (4*y3)), tmp4, xmask & ymask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (4, 4), (4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [inputs, flat_input], Original ATen: [aten.clone, aten.view]
stream0 = get_raw_stream(0)
triton_poi_fused_clone_view_0.run(primals_1, buf0, 64, 4, grid=grid(64, 4), stream=stream0)
buf1 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [matmul], Original ATen: [aten.mm]
extern_kernels.mm(buf0, reinterpret_tensor(primals_2, (4, 4), (1, 4), 0), out=buf1)
buf2 = buf1; del buf1 # reuse
# Topologically Sorted Source Nodes: [pow_1, sum_1, pow_2, sum_2, add, mul, distances], Original ATen: [aten.pow, aten.sum, aten.add, aten.mul, aten.sub]
triton_poi_fused_add_mul_pow_sub_sum_1.run(buf2, buf0, primals_2, 256, grid=grid(256), stream=stream0)
buf3 = empty_strided_cuda((64, ), (1, ), torch.int64)
# Topologically Sorted Source Nodes: [argmin], Original ATen: [aten.argmin]
triton_poi_fused_argmin_2.run(buf2, buf3, 64, grid=grid(64), stream=stream0)
buf4 = empty_strided_cpu((64, 1), (1, 1), torch.int64)
buf4.copy_(reinterpret_tensor(buf3, (64, 1), (1, 1), 0))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf5 = buf2; del buf2 # reuse
# Topologically Sorted Source Nodes: [scatter_], Original ATen: [aten.scatter]
triton_poi_fused_scatter_3.run(buf3, buf5, 256, grid=grid(256), stream=stream0)
del buf3
buf6 = buf0; del buf0 # reuse
# Topologically Sorted Source Nodes: [matmul_1], Original ATen: [aten.mm]
extern_kernels.mm(buf5, primals_2, out=buf6)
del primals_2
buf7 = empty_strided_cuda((), (), torch.float32)
buf12 = buf7; del buf7 # reuse
# Topologically Sorted Source Nodes: [inputs, e_latent_loss, mul_1, loss], Original ATen: [aten.clone, aten.mse_loss, aten.mul, aten.add]
triton_per_fused_add_clone_mse_loss_mul_4.run(buf12, buf6, primals_1, 1, 256, grid=grid(1), stream=stream0)
buf8 = empty_strided_cuda((4, ), (1, ), torch.float32)
# Topologically Sorted Source Nodes: [avg_probs], Original ATen: [aten.mean]
triton_per_fused_mean_5.run(buf5, buf8, 4, 64, grid=grid(4), stream=stream0)
buf9 = empty_strided_cuda((), (), torch.float32)
buf13 = buf9; del buf9 # reuse
# Topologically Sorted Source Nodes: [avg_probs, add_3, log, mul_2, sum_3, neg, perplexity], Original ATen: [aten.mean, aten.add, aten.log, aten.mul, aten.sum, aten.neg, aten.exp]
triton_per_fused_add_exp_log_mean_mul_neg_sum_6.run(buf13, buf8, 1, 4, grid=grid(1), stream=stream0)
del buf8
buf10 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [contiguous_1], Original ATen: [aten.clone]
triton_poi_fused_clone_7.run(primals_1, buf6, buf10, 16, 16, grid=grid(16, 16), stream=stream0)
buf11 = reinterpret_tensor(buf6, (4, 4, 4, 4), (64, 16, 4, 1), 0); del buf6 # reuse
# Topologically Sorted Source Nodes: [inputs, e_latent_loss], Original ATen: [aten.clone, aten.mse_loss, aten.mse_loss_backward]
triton_poi_fused_clone_mse_loss_mse_loss_backward_8.run(buf11, primals_1, 64, 4, grid=grid(64, 4), stream=stream0)
del primals_1
return (buf12, buf10, buf13, buf5, reinterpret_tensor(buf4, (), (), 0), buf11, reinterpret_tensor(buf5, (4, 64), (1, 4), 0), )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.nn.functional as F
class VectorQuantizer(nn.Module):
def __init__(self, num_embeddings, embedding_dim, commitment_cost):
super(VectorQuantizer, self).__init__()
self._embedding_dim = embedding_dim
self._num_embeddings = num_embeddings
self._embedding = nn.Embedding(self._num_embeddings, self.
_embedding_dim)
self._embedding.weight.data.uniform_(-1 / self._num_embeddings, 1 /
self._num_embeddings)
self._commitment_cost = commitment_cost
self.emb_indexes = []
def forward(self, inputs):
inputs = inputs.permute(0, 2, 3, 1).contiguous()
input_shape = inputs.shape
flat_input = inputs.view(-1, self._embedding_dim)
distances = torch.sum(flat_input ** 2, dim=1, keepdim=True
) + torch.sum(self._embedding.weight ** 2, dim=1
) - 2 * torch.matmul(flat_input, self._embedding.weight.t())
encoding_indices = torch.argmin(distances, dim=1).unsqueeze(1)
self.emb_indexes.extend(encoding_indices.cpu().detach().numpy()[0])
encodings = torch.zeros(encoding_indices.shape[0], self.
_num_embeddings, device=inputs.device)
encodings.scatter_(1, encoding_indices, 1)
quantized = torch.matmul(encodings, self._embedding.weight).view(
input_shape)
e_latent_loss = F.mse_loss(quantized.detach(), inputs)
q_latent_loss = F.mse_loss(quantized, inputs.detach())
loss = q_latent_loss + self._commitment_cost * e_latent_loss
quantized = inputs + (quantized - inputs).detach()
avg_probs = torch.mean(encodings, dim=0)
perplexity = torch.exp(-torch.sum(avg_probs * torch.log(avg_probs +
1e-10)))
return loss, quantized.permute(0, 3, 1, 2).contiguous(
), perplexity, encodings
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'num_embeddings': 4, 'embedding_dim': 4, 'commitment_cost': 4}
]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import math as tl_math
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_clone_view_0(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK:
tl.constexpr, XBLOCK: tl.constexpr):
ynumel = 64
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x1 = xindex
y0 = yindex
tmp0 = tl.load(in_ptr0 + (16 * x1 + 64 * (y0 // 16) + y0 % 16), xmask &
ymask, eviction_policy='evict_last')
tl.store(out_ptr0 + (x1 + 4 * y0), tmp0, xmask & ymask)
@triton.jit
def triton_poi_fused_add_mul_pow_sub_sum_1(in_out_ptr0, in_ptr0, in_ptr1,
xnumel, XBLOCK: tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = xindex // 4
x0 = xindex % 4
x2 = xindex
tmp0 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp8 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp11 = tl.load(in_ptr1 + 4 * x0, xmask, eviction_policy='evict_last')
tmp13 = tl.load(in_ptr1 + (1 + 4 * x0), xmask, eviction_policy='evict_last'
)
tmp16 = tl.load(in_ptr1 + (2 + 4 * x0), xmask, eviction_policy='evict_last'
)
tmp19 = tl.load(in_ptr1 + (3 + 4 * x0), xmask, eviction_policy='evict_last'
)
tmp23 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tmp0 * tmp0
tmp3 = tmp2 * tmp2
tmp4 = tmp1 + tmp3
tmp6 = tmp5 * tmp5
tmp7 = tmp4 + tmp6
tmp9 = tmp8 * tmp8
tmp10 = tmp7 + tmp9
tmp12 = tmp11 * tmp11
tmp14 = tmp13 * tmp13
tmp15 = tmp12 + tmp14
tmp17 = tmp16 * tmp16
tmp18 = tmp15 + tmp17
tmp20 = tmp19 * tmp19
tmp21 = tmp18 + tmp20
tmp22 = tmp10 + tmp21
tmp24 = 2.0
tmp25 = tmp23 * tmp24
tmp26 = tmp22 - tmp25
tl.store(in_out_ptr0 + x2, tmp26, xmask)
@triton.jit
def triton_poi_fused_argmin_2(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + 4 * x0, xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr0 + (1 + 4 * x0), xmask, eviction_policy='evict_last')
tmp17 = tl.load(in_ptr0 + (2 + 4 * x0), xmask, eviction_policy='evict_last'
)
tmp32 = tl.load(in_ptr0 + (3 + 4 * x0), xmask, eviction_policy='evict_last'
)
tmp2 = tmp0 < tmp1
tmp3 = tmp0 == tmp1
tmp4 = tmp0 != tmp0
tmp5 = tmp1 != tmp1
tmp6 = tmp4 > tmp5
tmp7 = tmp2 | tmp6
tmp8 = tmp4 & tmp5
tmp9 = tmp3 | tmp8
tmp10 = tl.full([1], 0, tl.int64)
tmp11 = tl.full([1], 1, tl.int64)
tmp12 = tmp10 < tmp11
tmp13 = tmp9 & tmp12
tmp14 = tmp7 | tmp13
tmp15 = tl.where(tmp14, tmp0, tmp1)
tmp16 = tl.where(tmp14, tmp10, tmp11)
tmp18 = tmp15 < tmp17
tmp19 = tmp15 == tmp17
tmp20 = tmp15 != tmp15
tmp21 = tmp17 != tmp17
tmp22 = tmp20 > tmp21
tmp23 = tmp18 | tmp22
tmp24 = tmp20 & tmp21
tmp25 = tmp19 | tmp24
tmp26 = tl.full([1], 2, tl.int64)
tmp27 = tmp16 < tmp26
tmp28 = tmp25 & tmp27
tmp29 = tmp23 | tmp28
tmp30 = tl.where(tmp29, tmp15, tmp17)
tmp31 = tl.where(tmp29, tmp16, tmp26)
tmp33 = tmp30 < tmp32
tmp34 = tmp30 == tmp32
tmp35 = tmp30 != tmp30
tmp36 = tmp32 != tmp32
tmp37 = tmp35 > tmp36
tmp38 = tmp33 | tmp37
tmp39 = tmp35 & tmp36
tmp40 = tmp34 | tmp39
tmp41 = tl.full([1], 3, tl.int64)
tmp42 = tmp31 < tmp41
tmp43 = tmp40 & tmp42
tmp44 = tmp38 | tmp43
tl.where(tmp44, tmp30, tmp32)
tmp46 = tl.where(tmp44, tmp31, tmp41)
tl.store(out_ptr0 + x0, tmp46, xmask)
@triton.jit
def triton_poi_fused_scatter_3(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = xindex // 4
x0 = xindex % 4
x2 = xindex
tmp0 = tl.load(in_ptr0 + x1, xmask, eviction_policy='evict_last')
tmp1 = x0
tmp2 = tmp0 == tmp1
tmp3 = 1.0
tmp4 = 0.0
tmp5 = tl.where(tmp2, tmp3, tmp4)
tl.store(out_ptr0 + x2, tmp5, xmask)
@triton.jit
def triton_per_fused_add_clone_mse_loss_mul_4(in_out_ptr0, in_ptr0, in_ptr1,
xnumel, rnumel):
XBLOCK: tl.constexpr = 1
RBLOCK: tl.constexpr = 256
xoffset = tl.program_id(0) * XBLOCK
tl.full([1], xoffset, tl.int32)
tl.full([RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[:]
tl.full([RBLOCK], True, tl.int1)
r3 = rindex
r0 = rindex % 4
r1 = rindex // 4 % 16
r2 = rindex // 64
tmp0 = tl.load(in_ptr0 + r3, None)
tmp1 = tl.load(in_ptr1 + (r1 + 16 * r0 + 64 * r2), None,
eviction_policy='evict_last')
tmp2 = tmp0 - tmp1
tmp3 = tmp2 * tmp2
tmp4 = tl.broadcast_to(tmp3, [RBLOCK])
tmp6 = triton_helpers.promote_to_tensor(tl.sum(tmp4, 0))
tmp7 = 256.0
tmp8 = tmp6 / tmp7
tmp9 = 4.0
tmp10 = tmp8 * tmp9
tmp11 = tmp8 + tmp10
tl.debug_barrier()
tl.store(in_out_ptr0 + tl.full([1], 0, tl.int32), tmp11, None)
@triton.jit
def triton_per_fused_mean_5(in_ptr0, out_ptr0, xnumel, rnumel, XBLOCK: tl.
constexpr):
xnumel = 4
RBLOCK: tl.constexpr = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
r1 = rindex
x0 = xindex
tmp0 = tl.load(in_ptr0 + (x0 + 4 * r1), xmask, other=0.0)
tmp1 = tl.broadcast_to(tmp0, [XBLOCK, RBLOCK])
tmp3 = tl.where(xmask, tmp1, 0)
tmp4 = tl.sum(tmp3, 1)[:, None]
tl.store(out_ptr0 + x0, tmp4, xmask)
@triton.jit
def triton_per_fused_add_exp_log_mean_mul_neg_sum_6(in_out_ptr0, in_ptr0,
xnumel, rnumel, XBLOCK: tl.constexpr):
RBLOCK: tl.constexpr = 4
xoffset = tl.program_id(0) * XBLOCK
xoffset + tl.arange(0, XBLOCK)[:, None]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[None, :]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
r0 = rindex
tmp0 = tl.load(in_ptr0 + r0, None)
tmp1 = 64.0
tmp2 = tmp0 / tmp1
tmp3 = 1e-10
tmp4 = tmp2 + tmp3
tmp5 = tl_math.log(tmp4)
tmp6 = tmp2 * tmp5
tmp7 = tl.broadcast_to(tmp6, [XBLOCK, RBLOCK])
tmp9 = tl.sum(tmp7, 1)[:, None]
tmp10 = -tmp9
tmp11 = tl_math.exp(tmp10)
tl.debug_barrier()
tl.store(in_out_ptr0 + tl.full([XBLOCK, 1], 0, tl.int32), tmp11, None)
@triton.jit
def triton_poi_fused_clone_7(in_ptr0, in_ptr1, out_ptr0, ynumel, xnumel,
YBLOCK: tl.constexpr, XBLOCK: tl.constexpr):
ynumel = 16
xnumel = 16
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y3 = yindex
y0 = yindex % 4
y1 = yindex // 4
tmp0 = tl.load(in_ptr0 + (x2 + 16 * y3), xmask & ymask)
tmp1 = tl.load(in_ptr1 + (y0 + 4 * x2 + 64 * y1), xmask & ymask)
tmp2 = tmp1 - tmp0
tmp3 = tmp0 + tmp2
tl.store(out_ptr0 + (x2 + 16 * y3), tmp3, xmask & ymask)
@triton.jit
def triton_poi_fused_clone_mse_loss_mse_loss_backward_8(in_out_ptr0,
in_ptr0, ynumel, xnumel, YBLOCK: tl.constexpr, XBLOCK: tl.constexpr):
ynumel = 64
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y3 = yindex
y0 = yindex % 16
y1 = yindex // 16
tmp0 = tl.load(in_out_ptr0 + (x2 + 4 * y3), xmask & ymask,
eviction_policy='evict_last')
tmp1 = tl.load(in_ptr0 + (y0 + 16 * x2 + 64 * y1), xmask & ymask,
eviction_policy='evict_last')
tmp2 = tmp0 - tmp1
tmp3 = 0.0078125
tmp4 = tmp2 * tmp3
tl.debug_barrier()
tl.store(in_out_ptr0 + (x2 + 4 * y3), tmp4, xmask & ymask)
def call(args):
primals_1, primals_2 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (4, 4), (4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_clone_view_0[grid(64, 4)](primals_1, buf0, 64, 4,
XBLOCK=4, YBLOCK=32, num_warps=4, num_stages=1)
buf1 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
extern_kernels.mm(buf0, reinterpret_tensor(primals_2, (4, 4), (1, 4
), 0), out=buf1)
buf2 = buf1
del buf1
triton_poi_fused_add_mul_pow_sub_sum_1[grid(256)](buf2, buf0,
primals_2, 256, XBLOCK=256, num_warps=4, num_stages=1)
buf3 = empty_strided_cuda((64,), (1,), torch.int64)
triton_poi_fused_argmin_2[grid(64)](buf2, buf3, 64, XBLOCK=64,
num_warps=1, num_stages=1)
buf4 = empty_strided_cpu((64, 1), (1, 1), torch.int64)
buf4.copy_(reinterpret_tensor(buf3, (64, 1), (1, 1), 0))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf5 = buf2
del buf2
triton_poi_fused_scatter_3[grid(256)](buf3, buf5, 256, XBLOCK=256,
num_warps=4, num_stages=1)
del buf3
buf6 = buf0
del buf0
extern_kernels.mm(buf5, primals_2, out=buf6)
del primals_2
buf7 = empty_strided_cuda((), (), torch.float32)
buf12 = buf7
del buf7
triton_per_fused_add_clone_mse_loss_mul_4[grid(1)](buf12, buf6,
primals_1, 1, 256, num_warps=2, num_stages=1)
buf8 = empty_strided_cuda((4,), (1,), torch.float32)
triton_per_fused_mean_5[grid(4)](buf5, buf8, 4, 64, XBLOCK=1,
num_warps=2, num_stages=1)
buf9 = empty_strided_cuda((), (), torch.float32)
buf13 = buf9
del buf9
triton_per_fused_add_exp_log_mean_mul_neg_sum_6[grid(1)](buf13,
buf8, 1, 4, XBLOCK=1, num_warps=2, num_stages=1)
del buf8
buf10 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
triton_poi_fused_clone_7[grid(16, 16)](primals_1, buf6, buf10, 16,
16, XBLOCK=16, YBLOCK=16, num_warps=4, num_stages=1)
buf11 = reinterpret_tensor(buf6, (4, 4, 4, 4), (64, 16, 4, 1), 0)
del buf6
triton_poi_fused_clone_mse_loss_mse_loss_backward_8[grid(64, 4)](buf11,
primals_1, 64, 4, XBLOCK=4, YBLOCK=32, num_warps=4, num_stages=1)
del primals_1
return buf12, buf10, buf13, buf5, reinterpret_tensor(buf4, (), (), 0
), buf11, reinterpret_tensor(buf5, (4, 64), (1, 4), 0)
class VectorQuantizerNew(nn.Module):
def __init__(self, num_embeddings, embedding_dim, commitment_cost):
super(VectorQuantizerNew, self).__init__()
self._embedding_dim = embedding_dim
self._num_embeddings = num_embeddings
self._embedding = nn.Embedding(self._num_embeddings, self.
_embedding_dim)
self._embedding.weight.data.uniform_(-1 / self._num_embeddings, 1 /
self._num_embeddings)
self._commitment_cost = commitment_cost
self.emb_indexes = []
def forward(self, input_0):
primals_2 = self._embedding.weight
primals_1 = input_0
output = call([primals_1, primals_2])
return output[0], output[1], output[2], output[3]
| imatge-upc/pixelcoordEDL | VectorQuantizer | false | 6,904 | [
"MIT"
] | 1 | 353632feed6ac8c93758c1a2a1b7a477e7ff053c | https://github.com/imatge-upc/pixelcoordEDL/tree/353632feed6ac8c93758c1a2a1b7a477e7ff053c | import torch
import torch.nn as nn
import torch.nn.functional as F
class Model(nn.Module):
def __init__(self, num_embeddings, embedding_dim, commitment_cost):
super().__init__()
self._embedding_dim = embedding_dim
self._num_embeddings = num_embeddings
self._embedding = nn.Embedding(self._num_embeddings, self.
_embedding_dim)
self._embedding.weight.data.uniform_(-1 / self._num_embeddings, 1 /
self._num_embeddings)
self._commitment_cost = commitment_cost
self.emb_indexes = []
def forward(self, inputs):
inputs = inputs.permute(0, 2, 3, 1).contiguous()
input_shape = inputs.shape
flat_input = inputs.view(-1, self._embedding_dim)
distances = torch.sum(flat_input ** 2, dim=1, keepdim=True
) + torch.sum(self._embedding.weight ** 2, dim=1
) - 2 * torch.matmul(flat_input, self._embedding.weight.t())
encoding_indices = torch.argmin(distances, dim=1).unsqueeze(1)
self.emb_indexes.extend(encoding_indices.cpu().detach().numpy()[0])
encodings = torch.zeros(encoding_indices.shape[0], self.
_num_embeddings, device=inputs.device)
encodings.scatter_(1, encoding_indices, 1)
quantized = torch.matmul(encodings, self._embedding.weight).view(
input_shape)
e_latent_loss = F.mse_loss(quantized.detach(), inputs)
q_latent_loss = F.mse_loss(quantized, inputs.detach())
loss = q_latent_loss + self._commitment_cost * e_latent_loss
quantized = inputs + (quantized - inputs).detach()
avg_probs = torch.mean(encodings, dim=0)
perplexity = torch.exp(-torch.sum(avg_probs * torch.log(avg_probs +
1e-10)))
return loss, quantized.permute(0, 3, 1, 2).contiguous(
), perplexity, encodings
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'num_embeddings': 4, 'embedding_dim': 4, 'commitment_cost': 4}
]
|
NavigatorBranch | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/uo/cuobiyjbapmnhkmwnq4daan4m3vy2y3j6thhvhnvjnoqztnpy23o.py
# Unsorted Source Nodes: [], Original ATen: []
# Source node to ATen node mapping:
triton_poi_fused_0 = async_compile.triton('triton_poi_fused_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[512, 16], tile_hint=TileHint.SQUARE,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_0(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 512
xnumel = 9
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y3 = yindex
y0 = yindex % 4
y1 = (yindex // 4)
tmp0 = tl.load(in_ptr0 + (x2 + (9*y3)), xmask & ymask, eviction_policy='evict_last')
tl.store(out_ptr0 + (y0 + (4*x2) + (36*y1)), tmp0, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/j5/cj5nf2owtsdm2zwcezqxpyn63iwddjyadpotkhm2ua52inoqxdcl.py
# Unsorted Source Nodes: [], Original ATen: []
# Source node to ATen node mapping:
triton_poi_fused_1 = async_compile.triton('triton_poi_fused_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16, 16], tile_hint=TileHint.SQUARE,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_1(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 16
xnumel = 16
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y3 = yindex
y0 = yindex % 4
y1 = (yindex // 4)
tmp0 = tl.load(in_ptr0 + (x2 + (16*y3)), xmask & ymask)
tl.store(out_ptr0 + (y0 + (4*x2) + (64*y1)), tmp0, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/gg/cgg6qm3q5dkubb7doxzvdjholdaslzdurfqpcxkw2o6qtw7fmkwz.py
# Topologically Sorted Source Nodes: [y, y_1, z], Original ATen: [aten.convolution, aten.relu]
# Source node to ATen node mapping:
# y => convolution
# y_1 => relu
# z => convolution_1
# Graph fragment:
# %convolution : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%primals_3, %primals_1, %primals_2, [1, 1], [1, 1], [1, 1], False, [0, 0], 1), kwargs = {})
# %relu : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%convolution,), kwargs = {})
# %convolution_1 : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%relu, %primals_4, %primals_5, [1, 1], [0, 0], [1, 1], False, [0, 0], 1), kwargs = {})
triton_poi_fused_convolution_relu_2 = async_compile.triton('triton_poi_fused_convolution_relu_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[512, 16], tile_hint=TileHint.DEFAULT,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32', 5: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_relu_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_relu_2(in_ptr0, in_ptr1, out_ptr0, out_ptr1, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 512
xnumel = 16
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 128
y1 = (yindex // 128)
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + (128*x2) + (2048*y1)), xmask & ymask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (y0), ymask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1, 1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tl.store(out_ptr0 + (x2 + (16*y3)), tmp4, xmask & ymask)
tl.store(out_ptr1 + (y0 + (128*x2) + (2048*y1)), tmp4, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/ur/curff7s5hkywtnymj6ypek5datxmdsklderihqz6yrgsqpmvwflf.py
# Topologically Sorted Source Nodes: [z], Original ATen: [aten.convolution]
# Source node to ATen node mapping:
# z => convolution_1
# Graph fragment:
# %convolution_1 : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%relu, %primals_4, %primals_5, [1, 1], [0, 0], [1, 1], False, [0, 0], 1), kwargs = {})
triton_poi_fused_convolution_3 = async_compile.triton('triton_poi_fused_convolution_3', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16, 16], tile_hint=TileHint.DEFAULT,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_3', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_3(in_ptr0, in_ptr1, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 16
xnumel = 16
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = (yindex // 4)
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + (4*x2) + (64*y1)), xmask & ymask)
tmp1 = tl.load(in_ptr1 + (y0), ymask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(out_ptr0 + (x2 + (16*y3)), tmp2, xmask & ymask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5 = args
args.clear()
assert_size_stride(primals_1, (128, 4, 3, 3), (36, 9, 3, 1))
assert_size_stride(primals_2, (128, ), (1, ))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_4, (4, 128, 1, 1), (128, 1, 1, 1))
assert_size_stride(primals_5, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((128, 4, 3, 3), (36, 1, 12, 4), torch.float32)
# Unsorted Source Nodes: [], Original ATen: []
stream0 = get_raw_stream(0)
triton_poi_fused_0.run(primals_1, buf0, 512, 9, grid=grid(512, 9), stream=stream0)
del primals_1
buf1 = empty_strided_cuda((4, 4, 4, 4), (64, 1, 16, 4), torch.float32)
# Unsorted Source Nodes: [], Original ATen: []
triton_poi_fused_1.run(primals_3, buf1, 16, 16, grid=grid(16, 16), stream=stream0)
del primals_3
# Topologically Sorted Source Nodes: [y], Original ATen: [aten.convolution]
buf2 = extern_kernels.convolution(buf1, buf0, stride=(1, 1), padding=(1, 1), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf2, (4, 128, 4, 4), (2048, 1, 512, 128))
buf3 = empty_strided_cuda((4, 128, 4, 4), (2048, 16, 4, 1), torch.float32)
buf4 = empty_strided_cuda((4, 128, 4, 4), (2048, 1, 512, 128), torch.float32)
# Topologically Sorted Source Nodes: [y, y_1, z], Original ATen: [aten.convolution, aten.relu]
triton_poi_fused_convolution_relu_2.run(buf2, primals_2, buf3, buf4, 512, 16, grid=grid(512, 16), stream=stream0)
del buf2
del primals_2
# Topologically Sorted Source Nodes: [z], Original ATen: [aten.convolution]
buf5 = extern_kernels.convolution(buf4, primals_4, stride=(1, 1), padding=(0, 0), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf5, (4, 4, 4, 4), (64, 1, 16, 4))
del buf4
buf6 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [z], Original ATen: [aten.convolution]
triton_poi_fused_convolution_3.run(buf5, primals_5, buf6, 16, 16, grid=grid(16, 16), stream=stream0)
del buf5
del primals_5
return (reinterpret_tensor(buf6, (4, 64), (64, 1), 0), buf3, buf0, buf1, primals_4, buf3, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((128, 4, 3, 3), (36, 9, 3, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((128, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, 128, 1, 1), (128, 1, 1, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
def conv1x1(in_channels, out_channels, stride=1, groups=1, bias=False):
"""
Convolution 1x1 layer.
Parameters:
----------
in_channels : int
Number of input channels.
out_channels : int
Number of output channels.
stride : int or tuple/list of 2 int, default 1
Strides of the convolution.
groups : int, default 1
Number of groups.
bias : bool, default False
Whether the layer uses a bias vector.
"""
return nn.Conv2d(in_channels=in_channels, out_channels=out_channels,
kernel_size=1, stride=stride, groups=groups, bias=bias)
def conv3x3(in_channels, out_channels, stride=1, padding=1, dilation=1,
groups=1, bias=False):
"""
Convolution 3x3 layer.
Parameters:
----------
in_channels : int
Number of input channels.
out_channels : int
Number of output channels.
stride : int or tuple/list of 2 int, default 1
Strides of the convolution.
padding : int or tuple/list of 2 int, default 1
Padding value for convolution layer.
groups : int, default 1
Number of groups.
bias : bool, default False
Whether the layer uses a bias vector.
"""
return nn.Conv2d(in_channels=in_channels, out_channels=out_channels,
kernel_size=3, stride=stride, padding=padding, dilation=dilation,
groups=groups, bias=bias)
class Flatten(nn.Module):
"""
Simple flatten module.
"""
def forward(self, x):
return x.view(x.size(0), -1)
class NavigatorBranch(nn.Module):
"""
Navigator branch block for Navigator unit.
Parameters:
----------
in_channels : int
Number of input channels.
out_channels : int
Number of output channels.
stride : int or tuple/list of 2 int
Strides of the convolution.
"""
def __init__(self, in_channels, out_channels, stride):
super(NavigatorBranch, self).__init__()
mid_channels = 128
self.down_conv = conv3x3(in_channels=in_channels, out_channels=
mid_channels, stride=stride, bias=True)
self.activ = nn.ReLU(inplace=False)
self.tidy_conv = conv1x1(in_channels=mid_channels, out_channels=
out_channels, bias=True)
self.flatten = Flatten()
def forward(self, x):
y = self.down_conv(x)
y = self.activ(y)
z = self.tidy_conv(y)
z = self.flatten(z)
return z, y
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'in_channels': 4, 'out_channels': 4, 'stride': 1}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_0(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK: tl.
constexpr, XBLOCK: tl.constexpr):
ynumel = 512
xnumel = 9
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y3 = yindex
y0 = yindex % 4
y1 = yindex // 4
tmp0 = tl.load(in_ptr0 + (x2 + 9 * y3), xmask & ymask, eviction_policy=
'evict_last')
tl.store(out_ptr0 + (y0 + 4 * x2 + 36 * y1), tmp0, xmask & ymask)
@triton.jit
def triton_poi_fused_1(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK: tl.
constexpr, XBLOCK: tl.constexpr):
ynumel = 16
xnumel = 16
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y3 = yindex
y0 = yindex % 4
y1 = yindex // 4
tmp0 = tl.load(in_ptr0 + (x2 + 16 * y3), xmask & ymask)
tl.store(out_ptr0 + (y0 + 4 * x2 + 64 * y1), tmp0, xmask & ymask)
@triton.jit
def triton_poi_fused_convolution_relu_2(in_ptr0, in_ptr1, out_ptr0,
out_ptr1, ynumel, xnumel, YBLOCK: tl.constexpr, XBLOCK: tl.constexpr):
ynumel = 512
xnumel = 16
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 128
y1 = yindex // 128
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + 128 * x2 + 2048 * y1), xmask & ymask,
eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + y0, ymask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1, 1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tl.store(out_ptr0 + (x2 + 16 * y3), tmp4, xmask & ymask)
tl.store(out_ptr1 + (y0 + 128 * x2 + 2048 * y1), tmp4, xmask & ymask)
@triton.jit
def triton_poi_fused_convolution_3(in_ptr0, in_ptr1, out_ptr0, ynumel,
xnumel, YBLOCK: tl.constexpr, XBLOCK: tl.constexpr):
ynumel = 16
xnumel = 16
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = yindex // 4
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + 4 * x2 + 64 * y1), xmask & ymask)
tmp1 = tl.load(in_ptr1 + y0, ymask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(out_ptr0 + (x2 + 16 * y3), tmp2, xmask & ymask)
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5 = args
args.clear()
assert_size_stride(primals_1, (128, 4, 3, 3), (36, 9, 3, 1))
assert_size_stride(primals_2, (128,), (1,))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_4, (4, 128, 1, 1), (128, 1, 1, 1))
assert_size_stride(primals_5, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((128, 4, 3, 3), (36, 1, 12, 4), torch.float32
)
get_raw_stream(0)
triton_poi_fused_0[grid(512, 9)](primals_1, buf0, 512, 9, XBLOCK=16,
YBLOCK=64, num_warps=4, num_stages=1)
del primals_1
buf1 = empty_strided_cuda((4, 4, 4, 4), (64, 1, 16, 4), torch.float32)
triton_poi_fused_1[grid(16, 16)](primals_3, buf1, 16, 16, XBLOCK=16,
YBLOCK=16, num_warps=4, num_stages=1)
del primals_3
buf2 = extern_kernels.convolution(buf1, buf0, stride=(1, 1),
padding=(1, 1), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf2, (4, 128, 4, 4), (2048, 1, 512, 128))
buf3 = empty_strided_cuda((4, 128, 4, 4), (2048, 16, 4, 1), torch.
float32)
buf4 = empty_strided_cuda((4, 128, 4, 4), (2048, 1, 512, 128),
torch.float32)
triton_poi_fused_convolution_relu_2[grid(512, 16)](buf2, primals_2,
buf3, buf4, 512, 16, XBLOCK=16, YBLOCK=64, num_warps=4,
num_stages=1)
del buf2
del primals_2
buf5 = extern_kernels.convolution(buf4, primals_4, stride=(1, 1),
padding=(0, 0), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf5, (4, 4, 4, 4), (64, 1, 16, 4))
del buf4
buf6 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
triton_poi_fused_convolution_3[grid(16, 16)](buf5, primals_5, buf6,
16, 16, XBLOCK=16, YBLOCK=16, num_warps=4, num_stages=1)
del buf5
del primals_5
return reinterpret_tensor(buf6, (4, 64), (64, 1), 0
), buf3, buf0, buf1, primals_4, buf3
def conv1x1(in_channels, out_channels, stride=1, groups=1, bias=False):
"""
Convolution 1x1 layer.
Parameters:
----------
in_channels : int
Number of input channels.
out_channels : int
Number of output channels.
stride : int or tuple/list of 2 int, default 1
Strides of the convolution.
groups : int, default 1
Number of groups.
bias : bool, default False
Whether the layer uses a bias vector.
"""
return nn.Conv2d(in_channels=in_channels, out_channels=out_channels,
kernel_size=1, stride=stride, groups=groups, bias=bias)
def conv3x3(in_channels, out_channels, stride=1, padding=1, dilation=1,
groups=1, bias=False):
"""
Convolution 3x3 layer.
Parameters:
----------
in_channels : int
Number of input channels.
out_channels : int
Number of output channels.
stride : int or tuple/list of 2 int, default 1
Strides of the convolution.
padding : int or tuple/list of 2 int, default 1
Padding value for convolution layer.
groups : int, default 1
Number of groups.
bias : bool, default False
Whether the layer uses a bias vector.
"""
return nn.Conv2d(in_channels=in_channels, out_channels=out_channels,
kernel_size=3, stride=stride, padding=padding, dilation=dilation,
groups=groups, bias=bias)
class Flatten(nn.Module):
"""
Simple flatten module.
"""
def forward(self, x):
return x.view(x.size(0), -1)
class NavigatorBranchNew(nn.Module):
"""
Navigator branch block for Navigator unit.
Parameters:
----------
in_channels : int
Number of input channels.
out_channels : int
Number of output channels.
stride : int or tuple/list of 2 int
Strides of the convolution.
"""
def __init__(self, in_channels, out_channels, stride):
super(NavigatorBranchNew, self).__init__()
mid_channels = 128
self.down_conv = conv3x3(in_channels=in_channels, out_channels=
mid_channels, stride=stride, bias=True)
self.activ = nn.ReLU(inplace=False)
self.tidy_conv = conv1x1(in_channels=mid_channels, out_channels=
out_channels, bias=True)
self.flatten = Flatten()
def forward(self, input_0):
primals_1 = self.down_conv.weight
primals_2 = self.down_conv.bias
primals_4 = self.tidy_conv.weight
primals_5 = self.tidy_conv.bias
primals_3 = input_0
output = call([primals_1, primals_2, primals_3, primals_4, primals_5])
return output[0], output[1]
| iofthetiger/pkuad | NavigatorBranch | false | 6,905 | [
"Apache-2.0"
] | 1 | 07496d108c614c84be028f344830becc9cac8fe5 | https://github.com/iofthetiger/pkuad/tree/07496d108c614c84be028f344830becc9cac8fe5 | import torch
import torch.nn as nn
def conv1x1(in_channels, out_channels, stride=1, groups=1, bias=False):
"""
Convolution 1x1 layer.
Parameters:
----------
in_channels : int
Number of input channels.
out_channels : int
Number of output channels.
stride : int or tuple/list of 2 int, default 1
Strides of the convolution.
groups : int, default 1
Number of groups.
bias : bool, default False
Whether the layer uses a bias vector.
"""
return nn.Conv2d(in_channels=in_channels, out_channels=out_channels,
kernel_size=1, stride=stride, groups=groups, bias=bias)
def conv3x3(in_channels, out_channels, stride=1, padding=1, dilation=1,
groups=1, bias=False):
"""
Convolution 3x3 layer.
Parameters:
----------
in_channels : int
Number of input channels.
out_channels : int
Number of output channels.
stride : int or tuple/list of 2 int, default 1
Strides of the convolution.
padding : int or tuple/list of 2 int, default 1
Padding value for convolution layer.
groups : int, default 1
Number of groups.
bias : bool, default False
Whether the layer uses a bias vector.
"""
return nn.Conv2d(in_channels=in_channels, out_channels=out_channels,
kernel_size=3, stride=stride, padding=padding, dilation=dilation,
groups=groups, bias=bias)
class Flatten(nn.Module):
"""
Simple flatten module.
"""
def forward(self, x):
return x.view(x.size(0), -1)
class Model(nn.Module):
"""
Navigator branch block for Navigator unit.
Parameters:
----------
in_channels : int
Number of input channels.
out_channels : int
Number of output channels.
stride : int or tuple/list of 2 int
Strides of the convolution.
"""
def __init__(self, in_channels, out_channels, stride):
super().__init__()
mid_channels = 128
self.down_conv = conv3x3(in_channels=in_channels, out_channels=
mid_channels, stride=stride, bias=True)
self.activ = nn.ReLU(inplace=False)
self.tidy_conv = conv1x1(in_channels=mid_channels, out_channels=
out_channels, bias=True)
self.flatten = Flatten()
def forward(self, x):
y = self.down_conv(x)
y = self.activ(y)
z = self.tidy_conv(y)
z = self.flatten(z)
return z, y
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4, 4, 1]
|
SCLN | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/ah/cahvhw7w2mv7l7skir3lhl2g7tllk2lyu3j3qszf4szy4m4elkyn.py
# Topologically Sorted Source Nodes: [mu, sigma, sub, add, y, mul, o], Original ATen: [aten.mean, aten.std, aten.sub, aten.add, aten.div, aten.mul]
# Source node to ATen node mapping:
# add => add
# mu => mean
# mul => mul
# o => add_1
# sigma => sqrt, var
# sub => sub
# y => div
# Graph fragment:
# %mean : [num_users=1] = call_function[target=torch.ops.aten.mean.dim](args = (%primals_1, [-1], True), kwargs = {})
# %var : [num_users=1] = call_function[target=torch.ops.aten.var.correction](args = (%primals_1, [-1]), kwargs = {correction: 1.0, keepdim: True})
# %sqrt : [num_users=1] = call_function[target=torch.ops.aten.sqrt.default](args = (%var,), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%primals_1, %mean), kwargs = {})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%sqrt, 1e-08), kwargs = {})
# %div : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%sub, %add), kwargs = {})
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%getitem_1, %div), kwargs = {})
# %add_1 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul, %getitem), kwargs = {})
triton_poi_fused_add_div_mean_mul_std_sub_0 = async_compile.triton('triton_poi_fused_add_div_mean_mul_std_sub_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_div_mean_mul_std_sub_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 7, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_div_mean_mul_std_sub_0(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 4
x1 = (xindex // 4)
x2 = xindex
tmp0 = tl.load(in_ptr0 + (4 + x0 + (8*x1)), xmask)
tmp1 = tl.load(in_ptr1 + (x2), xmask)
tmp2 = tl.load(in_ptr1 + (4*x1), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr1 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr1 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr1 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp30 = tl.load(in_ptr0 + (x0 + (8*x1)), xmask)
tmp4 = tmp2 + tmp3
tmp6 = tmp4 + tmp5
tmp8 = tmp6 + tmp7
tmp9 = 4.0
tmp10 = tmp8 / tmp9
tmp11 = tmp1 - tmp10
tmp12 = tmp2 - tmp10
tmp13 = tmp12 * tmp12
tmp14 = tmp3 - tmp10
tmp15 = tmp14 * tmp14
tmp16 = tmp13 + tmp15
tmp17 = tmp5 - tmp10
tmp18 = tmp17 * tmp17
tmp19 = tmp16 + tmp18
tmp20 = tmp7 - tmp10
tmp21 = tmp20 * tmp20
tmp22 = tmp19 + tmp21
tmp23 = 3.0
tmp24 = tmp22 / tmp23
tmp25 = libdevice.sqrt(tmp24)
tmp26 = 1e-08
tmp27 = tmp25 + tmp26
tmp28 = tmp11 / tmp27
tmp29 = tmp0 * tmp28
tmp31 = tmp29 + tmp30
tl.store(out_ptr0 + (x2), tmp31, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (8, 4), (4, 1))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 8), (8, 1), torch.float32)
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.mm]
extern_kernels.mm(reinterpret_tensor(primals_3, (64, 4), (4, 1), 0), reinterpret_tensor(primals_2, (4, 8), (1, 4), 0), out=buf0)
del primals_2
buf1 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [mu, sigma, sub, add, y, mul, o], Original ATen: [aten.mean, aten.std, aten.sub, aten.add, aten.div, aten.mul]
stream0 = get_raw_stream(0)
triton_poi_fused_add_div_mean_mul_std_sub_0.run(buf0, primals_1, buf1, 256, grid=grid(256), stream=stream0)
del buf0
return (buf1, primals_1, reinterpret_tensor(primals_3, (64, 4), (4, 1), 0), )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((8, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
class LinearNorm(nn.Module):
""" LinearNorm Projection """
def __init__(self, in_features, out_features, bias=False):
super(LinearNorm, self).__init__()
self.linear = nn.Linear(in_features, out_features, bias)
nn.init.xavier_uniform_(self.linear.weight)
if bias:
nn.init.constant_(self.linear.bias, 0.0)
def forward(self, x):
x = self.linear(x)
return x
class SCLN(nn.Module):
""" Speaker Condition Layer Normalization """
def __init__(self, s_size, hidden_size, eps=1e-08, bias=False):
super(SCLN, self).__init__()
self.hidden_size = hidden_size
self.affine_layer = LinearNorm(s_size, 2 * hidden_size, bias)
self.eps = eps
def forward(self, x, s):
mu, sigma = torch.mean(x, dim=-1, keepdim=True), torch.std(x, dim=-
1, keepdim=True)
y = (x - mu) / (sigma + self.eps)
b, g = torch.split(self.affine_layer(s), self.hidden_size, dim=-1)
o = g * y + b
return o
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'s_size': 4, 'hidden_size': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime.triton_helpers import libdevice
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_add_div_mean_mul_std_sub_0(in_ptr0, in_ptr1, out_ptr0,
xnumel, XBLOCK: tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 4
x1 = xindex // 4
x2 = xindex
tmp0 = tl.load(in_ptr0 + (4 + x0 + 8 * x1), xmask)
tmp1 = tl.load(in_ptr1 + x2, xmask)
tmp2 = tl.load(in_ptr1 + 4 * x1, xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr1 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr1 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp7 = tl.load(in_ptr1 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp30 = tl.load(in_ptr0 + (x0 + 8 * x1), xmask)
tmp4 = tmp2 + tmp3
tmp6 = tmp4 + tmp5
tmp8 = tmp6 + tmp7
tmp9 = 4.0
tmp10 = tmp8 / tmp9
tmp11 = tmp1 - tmp10
tmp12 = tmp2 - tmp10
tmp13 = tmp12 * tmp12
tmp14 = tmp3 - tmp10
tmp15 = tmp14 * tmp14
tmp16 = tmp13 + tmp15
tmp17 = tmp5 - tmp10
tmp18 = tmp17 * tmp17
tmp19 = tmp16 + tmp18
tmp20 = tmp7 - tmp10
tmp21 = tmp20 * tmp20
tmp22 = tmp19 + tmp21
tmp23 = 3.0
tmp24 = tmp22 / tmp23
tmp25 = libdevice.sqrt(tmp24)
tmp26 = 1e-08
tmp27 = tmp25 + tmp26
tmp28 = tmp11 / tmp27
tmp29 = tmp0 * tmp28
tmp31 = tmp29 + tmp30
tl.store(out_ptr0 + x2, tmp31, xmask)
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (8, 4), (4, 1))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 8), (8, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_3, (64, 4), (4, 1), 0),
reinterpret_tensor(primals_2, (4, 8), (1, 4), 0), out=buf0)
del primals_2
buf1 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_add_div_mean_mul_std_sub_0[grid(256)](buf0,
primals_1, buf1, 256, XBLOCK=128, num_warps=4, num_stages=1)
del buf0
return buf1, primals_1, reinterpret_tensor(primals_3, (64, 4), (4, 1), 0)
class LinearNorm(nn.Module):
""" LinearNorm Projection """
def __init__(self, in_features, out_features, bias=False):
super(LinearNorm, self).__init__()
self.linear = nn.Linear(in_features, out_features, bias)
nn.init.xavier_uniform_(self.linear.weight)
if bias:
nn.init.constant_(self.linear.bias, 0.0)
def forward(self, x):
x = self.linear(x)
return x
class SCLNNew(nn.Module):
""" Speaker Condition Layer Normalization """
def __init__(self, s_size, hidden_size, eps=1e-08, bias=False):
super(SCLNNew, self).__init__()
self.hidden_size = hidden_size
self.affine_layer = LinearNorm(s_size, 2 * hidden_size, bias)
self.eps = eps
def forward(self, input_0, input_1):
primals_2 = self.affine_layer.linear.weight
primals_1 = input_0
primals_3 = input_1
output = call([primals_1, primals_2, primals_3])
return output[0]
| ishine/Cross-Speaker-Emotion-Transfer | SCLN | false | 6,906 | [
"MIT"
] | 1 | 9d38e8058f5abc06167bac244d8ace083e2a6220 | https://github.com/ishine/Cross-Speaker-Emotion-Transfer/tree/9d38e8058f5abc06167bac244d8ace083e2a6220 | import torch
import torch.nn as nn
class LinearNorm(nn.Module):
""" LinearNorm Projection """
def __init__(self, in_features, out_features, bias=False):
super().__init__()
self.linear = nn.Linear(in_features, out_features, bias)
nn.init.xavier_uniform_(self.linear.weight)
if bias:
nn.init.constant_(self.linear.bias, 0.0)
def forward(self, x):
x = self.linear(x)
return x
class Model(nn.Module):
""" Speaker Condition Layer Normalization """
def __init__(self, s_size, hidden_size, eps=1e-08, bias=False):
super().__init__()
self.hidden_size = hidden_size
self.affine_layer = LinearNorm(s_size, 2 * hidden_size, bias)
self.eps = eps
def forward(self, x, s):
mu, sigma = torch.mean(x, dim=-1, keepdim=True), torch.std(x, dim=-
1, keepdim=True)
y = (x - mu) / (sigma + self.eps)
b, g = torch.split(self.affine_layer(s), self.hidden_size, dim=-1)
o = g * y + b
return o
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4, 4]
|
Tile | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/k3/ck3gmopv7gjt5rihteo5afwbo6re2c3gf53eqqxhkhqh25dwpw5i.py
# Topologically Sorted Source Nodes: [repeat], Original ATen: [aten.repeat]
# Source node to ATen node mapping:
# repeat => repeat
# Graph fragment:
# %repeat : [num_users=1] = call_function[target=torch.ops.aten.repeat.default](args = (%arg0_1, [1, 1, 1, 1]), kwargs = {})
triton_poi_fused_repeat_0 = async_compile.triton('triton_poi_fused_repeat_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_repeat_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_repeat_0(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (x0), xmask)
tl.store(out_ptr0 + (x0), tmp0, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [repeat], Original ATen: [aten.repeat]
stream0 = get_raw_stream(0)
triton_poi_fused_repeat_0.run(arg0_1, buf0, 256, grid=grid(256), stream=stream0)
del arg0_1
return (buf0, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([arg0_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
class Tile(nn.Module):
def __init__(self, max_size, dim):
super(Tile, self).__init__()
self.max_size = max_size
self.dim = dim
def forward(self, input):
return input.repeat(*[(self.max_size if x == self.dim else 1) for x in
range(len(input.shape))])
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'max_size': 4, 'dim': 4}]
| import torch
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_poi_fused_repeat_0(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + x0, xmask)
tl.store(out_ptr0 + x0, tmp0, xmask)
def call(args):
arg0_1, = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_repeat_0[grid(256)](arg0_1, buf0, 256, XBLOCK=128,
num_warps=4, num_stages=1)
del arg0_1
return buf0,
class TileNew(nn.Module):
def __init__(self, max_size, dim):
super(TileNew, self).__init__()
self.max_size = max_size
self.dim = dim
def forward(self, input_0):
arg0_1 = input_0
output = call([arg0_1])
return output[0]
| ixaxaar/pytorch-npi | Tile | false | 6,907 | [
"MIT"
] | 1 | 50b028840c00f7807fb6490ce6bb0918832dc360 | https://github.com/ixaxaar/pytorch-npi/tree/50b028840c00f7807fb6490ce6bb0918832dc360 | import torch
import torch.nn as nn
class Model(nn.Module):
def __init__(self, max_size, dim):
super().__init__()
self.max_size = max_size
self.dim = dim
def forward(self, input):
return input.repeat(*[(self.max_size if x == self.dim else 1) for x in
range(len(input.shape))])
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4, 4]
|
GlobalAvgPool2d | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/l3/cl35tzbhrd24dhunkbb6gjs54aklpyr46oikqhoylcgmkcmhujil.py
# Topologically Sorted Source Nodes: [inputs], Original ATen: [aten.mean]
# Source node to ATen node mapping:
# inputs => mean
# Graph fragment:
# %mean : [num_users=1] = call_function[target=torch.ops.aten.mean.dim](args = (%view, [2]), kwargs = {})
triton_per_fused_mean_0 = async_compile.triton('triton_per_fused_mean_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[16, 16],
reduction_hint=ReductionHint.INNER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused_mean_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 1, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused_mean_0(in_out_ptr0, in_ptr0, xnumel, rnumel, XBLOCK : tl.constexpr):
xnumel = 16
rnumel = 16
RBLOCK: tl.constexpr = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
roffset = 0
rmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
r1 = rindex
x0 = xindex
tmp0 = tl.load(in_ptr0 + (r1 + (16*x0)), xmask, other=0.0)
tmp1 = tl.broadcast_to(tmp0, [XBLOCK, RBLOCK])
tmp3 = tl.where(xmask, tmp1, 0)
tmp4 = tl.sum(tmp3, 1)[:, None]
tmp5 = 16.0
tmp6 = tmp4 / tmp5
tl.debug_barrier()
tl.store(in_out_ptr0 + (x0), tmp6, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
buf1 = buf0; del buf0 # reuse
# Topologically Sorted Source Nodes: [inputs], Original ATen: [aten.mean]
stream0 = get_raw_stream(0)
triton_per_fused_mean_0.run(buf1, arg0_1, 16, 16, grid=grid(16), stream=stream0)
del arg0_1
return (reinterpret_tensor(buf1, (4, 4, 1, 1), (4, 1, 1, 1), 0), )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([arg0_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.utils
class GlobalAvgPool2d(nn.Module):
def __init__(self):
"""Global average pooling over the input's spatial dimensions"""
super(GlobalAvgPool2d, self).__init__()
def forward(self, inputs):
in_size = inputs.size()
inputs = inputs.view((in_size[0], in_size[1], -1)).mean(dim=2)
inputs = inputs.view(in_size[0], in_size[1], 1, 1)
return inputs
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {}]
| import torch
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
import torch.nn as nn
import torch.utils
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_per_fused_mean_0(in_out_ptr0, in_ptr0, xnumel, rnumel, XBLOCK:
tl.constexpr):
xnumel = 16
RBLOCK: tl.constexpr = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
r1 = rindex
x0 = xindex
tmp0 = tl.load(in_ptr0 + (r1 + 16 * x0), xmask, other=0.0)
tmp1 = tl.broadcast_to(tmp0, [XBLOCK, RBLOCK])
tmp3 = tl.where(xmask, tmp1, 0)
tmp4 = tl.sum(tmp3, 1)[:, None]
tmp5 = 16.0
tmp6 = tmp4 / tmp5
tl.debug_barrier()
tl.store(in_out_ptr0 + x0, tmp6, xmask)
def call(args):
arg0_1, = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
buf1 = buf0
del buf0
get_raw_stream(0)
triton_per_fused_mean_0[grid(16)](buf1, arg0_1, 16, 16, XBLOCK=8,
num_warps=2, num_stages=1)
del arg0_1
return reinterpret_tensor(buf1, (4, 4, 1, 1), (4, 1, 1, 1), 0),
class GlobalAvgPool2dNew(nn.Module):
def __init__(self):
"""Global average pooling over the input's spatial dimensions"""
super(GlobalAvgPool2dNew, self).__init__()
def forward(self, input_0):
arg0_1 = input_0
output = call([arg0_1])
return output[0]
| jameslong95/FasterSeg | GlobalAvgPool2d | false | 6,908 | [
"MIT"
] | 1 | 872e04964ea46494a6018d9915cee5476e361c27 | https://github.com/jameslong95/FasterSeg/tree/872e04964ea46494a6018d9915cee5476e361c27 | import torch
import torch.nn as nn
import torch.utils
class Model(nn.Module):
def __init__(self):
"""Global average pooling over the input's spatial dimensions"""
super().__init__()
def forward(self, inputs):
in_size = inputs.size()
inputs = inputs.view((in_size[0], in_size[1], -1)).mean(dim=2)
inputs = inputs.view(in_size[0], in_size[1], 1, 1)
return inputs
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return []
|
Vec2ArousalNet | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/ai/caitnpldotnv4k4oj67wyecd2ig4qcjrnmr35rmx6o2vxx245xs3.py
# Topologically Sorted Source Nodes: [h], Original ATen: [aten.clamp, aten.ge]
# Source node to ATen node mapping:
# h => clamp_min
# Graph fragment:
# %clamp_min : [num_users=1] = call_function[target=torch.ops.aten.clamp_min.default](args = (%view_1, 0), kwargs = {})
# %ge_1 : [num_users=1] = call_function[target=torch.ops.aten.ge.Scalar](args = (%view_1, 0), kwargs = {})
triton_poi_fused_clamp_ge_0 = async_compile.triton('triton_poi_fused_clamp_ge_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*i1', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clamp_ge_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clamp_ge_0(in_ptr0, in_ptr1, out_ptr0, out_ptr1, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr1 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = 0.0
tmp4 = triton_helpers.maximum(tmp2, tmp3)
tmp5 = tmp2 >= tmp3
tl.store(out_ptr0 + (x2), tmp4, xmask)
tl.store(out_ptr1 + (x2), tmp5, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/ut/cutdhw2ppb7axwusafshofwwckhzhrvsrcxdtiduoohmfxybvmmi.py
# Topologically Sorted Source Nodes: [y], Original ATen: [aten.clamp, aten.ge, aten.le, aten.logical_and]
# Source node to ATen node mapping:
# y => clamp_max, clamp_min_1
# Graph fragment:
# %clamp_min_1 : [num_users=1] = call_function[target=torch.ops.aten.clamp_min.default](args = (%view_3, 0), kwargs = {})
# %clamp_max : [num_users=1] = call_function[target=torch.ops.aten.clamp_max.default](args = (%clamp_min_1, 1), kwargs = {})
# %ge : [num_users=1] = call_function[target=torch.ops.aten.ge.Scalar](args = (%view_3, 0), kwargs = {})
# %le : [num_users=1] = call_function[target=torch.ops.aten.le.Scalar](args = (%view_3, 1), kwargs = {})
# %logical_and : [num_users=1] = call_function[target=torch.ops.aten.logical_and.default](args = (%ge, %le), kwargs = {})
triton_poi_fused_clamp_ge_le_logical_and_1 = async_compile.triton('triton_poi_fused_clamp_ge_le_logical_and_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*i1', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clamp_ge_le_logical_and_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clamp_ge_le_logical_and_1(in_ptr0, in_ptr1, out_ptr0, out_ptr1, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr1 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = 0.0
tmp4 = triton_helpers.maximum(tmp2, tmp3)
tmp5 = 1.0
tmp6 = triton_helpers.minimum(tmp4, tmp5)
tmp7 = tmp2 >= tmp3
tmp8 = tmp2 <= tmp5
tmp9 = tmp7 & tmp8
tl.store(out_ptr0 + (x2), tmp6, xmask)
tl.store(out_ptr1 + (x2), tmp9, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5 = args
args.clear()
assert_size_stride(primals_1, (4, 4), (4, 1))
assert_size_stride(primals_2, (4, ), (1, ))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_4, (4, 4), (4, 1))
assert_size_stride(primals_5, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(primals_3, (64, 4), (4, 1), 0), reinterpret_tensor(primals_1, (4, 4), (1, 4), 0), out=buf0)
del primals_1
buf1 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
buf5 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.bool)
# Topologically Sorted Source Nodes: [h], Original ATen: [aten.clamp, aten.ge]
stream0 = get_raw_stream(0)
triton_poi_fused_clamp_ge_0.run(buf0, primals_2, buf1, buf5, 256, grid=grid(256), stream=stream0)
del primals_2
buf2 = buf0; del buf0 # reuse
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(buf1, (64, 4), (4, 1), 0), reinterpret_tensor(primals_4, (4, 4), (1, 4), 0), out=buf2)
buf3 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
buf4 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.bool)
# Topologically Sorted Source Nodes: [y], Original ATen: [aten.clamp, aten.ge, aten.le, aten.logical_and]
triton_poi_fused_clamp_ge_le_logical_and_1.run(buf2, primals_5, buf3, buf4, 256, grid=grid(256), stream=stream0)
del buf2
del primals_5
return (buf3, reinterpret_tensor(primals_3, (64, 4), (4, 1), 0), reinterpret_tensor(buf1, (64, 4), (4, 1), 0), buf4, primals_4, buf5, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.utils.data
class Vec2ArousalNet(torch.nn.Module):
def __init__(self, D_in, H, D_out):
super(Vec2ArousalNet, self).__init__()
self.layer_1 = torch.nn.Linear(D_in, H)
self.layer_2 = torch.nn.Linear(H, D_out)
def forward(self, x):
h = self.layer_1(x).clamp(min=0)
y = self.layer_2(h).clamp(min=0, max=1)
return y
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'D_in': 4, 'H': 4, 'D_out': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
import torch.utils.data
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_clamp_ge_0(in_ptr0, in_ptr1, out_ptr0, out_ptr1,
xnumel, XBLOCK: tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr1 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = 0.0
tmp4 = triton_helpers.maximum(tmp2, tmp3)
tmp5 = tmp2 >= tmp3
tl.store(out_ptr0 + x2, tmp4, xmask)
tl.store(out_ptr1 + x2, tmp5, xmask)
@triton.jit
def triton_poi_fused_clamp_ge_le_logical_and_1(in_ptr0, in_ptr1, out_ptr0,
out_ptr1, xnumel, XBLOCK: tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr1 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = 0.0
tmp4 = triton_helpers.maximum(tmp2, tmp3)
tmp5 = 1.0
tmp6 = triton_helpers.minimum(tmp4, tmp5)
tmp7 = tmp2 >= tmp3
tmp8 = tmp2 <= tmp5
tmp9 = tmp7 & tmp8
tl.store(out_ptr0 + x2, tmp6, xmask)
tl.store(out_ptr1 + x2, tmp9, xmask)
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5 = args
args.clear()
assert_size_stride(primals_1, (4, 4), (4, 1))
assert_size_stride(primals_2, (4,), (1,))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_4, (4, 4), (4, 1))
assert_size_stride(primals_5, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_3, (64, 4), (4, 1), 0),
reinterpret_tensor(primals_1, (4, 4), (1, 4), 0), out=buf0)
del primals_1
buf1 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
buf5 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.bool)
get_raw_stream(0)
triton_poi_fused_clamp_ge_0[grid(256)](buf0, primals_2, buf1, buf5,
256, XBLOCK=256, num_warps=4, num_stages=1)
del primals_2
buf2 = buf0
del buf0
extern_kernels.mm(reinterpret_tensor(buf1, (64, 4), (4, 1), 0),
reinterpret_tensor(primals_4, (4, 4), (1, 4), 0), out=buf2)
buf3 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
buf4 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.bool)
triton_poi_fused_clamp_ge_le_logical_and_1[grid(256)](buf2,
primals_5, buf3, buf4, 256, XBLOCK=128, num_warps=4, num_stages=1)
del buf2
del primals_5
return buf3, reinterpret_tensor(primals_3, (64, 4), (4, 1), 0
), reinterpret_tensor(buf1, (64, 4), (4, 1), 0), buf4, primals_4, buf5
class Vec2ArousalNetNew(torch.nn.Module):
def __init__(self, D_in, H, D_out):
super(Vec2ArousalNetNew, self).__init__()
self.layer_1 = torch.nn.Linear(D_in, H)
self.layer_2 = torch.nn.Linear(H, D_out)
def forward(self, input_0):
primals_1 = self.layer_1.weight
primals_2 = self.layer_1.bias
primals_4 = self.layer_2.weight
primals_5 = self.layer_2.bias
primals_3 = input_0
output = call([primals_1, primals_2, primals_3, primals_4, primals_5])
return output[0]
| jackvandrunen/hackuci18 | Vec2ArousalNet | false | 6,909 | [
"BSD-2-Clause"
] | 1 | fff3fd7d116a6a83f19229a17377b84922145ebd | https://github.com/jackvandrunen/hackuci18/tree/fff3fd7d116a6a83f19229a17377b84922145ebd | import torch
import torch.utils.data
class Model(torch.nn.Module):
def __init__(self, D_in, H, D_out):
super().__init__()
self.layer_1 = torch.nn.Linear(D_in, H)
self.layer_2 = torch.nn.Linear(H, D_out)
def forward(self, x):
h = self.layer_1(x).clamp(min=0)
y = self.layer_2(h).clamp(min=0, max=1)
return y
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4, 4, 4]
|
LinearAttention2d | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/zi/czioyfiql36jvbru3amu3iggyuvnn5c4pypwuaiss36muc2jqtqb.py
# Topologically Sorted Source Nodes: [add], Original ATen: [aten.add]
# Source node to ATen node mapping:
# add => add
# Graph fragment:
# %add : [num_users=2] = call_function[target=torch.ops.aten.add.Tensor](args = (%primals_1, %primals_2), kwargs = {})
triton_poi_fused_add_0 = async_compile.triton('triton_poi_fused_add_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_0(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (x0), xmask)
tmp1 = tl.load(in_ptr1 + (x0), xmask)
tmp2 = tmp0 + tmp1
tl.store(out_ptr0 + (x0), tmp2, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/sk/cskcg4lhauu2iopcbfiq3dbajlyfau73b5coowr5uwohijb36k3c.py
# Topologically Sorted Source Nodes: [softmax], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# softmax => amax, div, exp, sub, sum_1
# Graph fragment:
# %amax : [num_users=1] = call_function[target=torch.ops.aten.amax.default](args = (%view, [2], True), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%view, %amax), kwargs = {})
# %exp : [num_users=2] = call_function[target=torch.ops.aten.exp.default](args = (%sub,), kwargs = {})
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%exp, [2], True), kwargs = {})
# %div : [num_users=2] = call_function[target=torch.ops.aten.div.Tensor](args = (%exp, %sum_1), kwargs = {})
triton_per_fused__softmax_1 = async_compile.triton('triton_per_fused__softmax_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[4, 16],
reduction_hint=ReductionHint.INNER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused__softmax_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 2, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused__softmax_1(in_ptr0, out_ptr2, xnumel, rnumel, XBLOCK : tl.constexpr):
xnumel = 4
rnumel = 16
RBLOCK: tl.constexpr = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
roffset = 0
rmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
r1 = rindex
x0 = xindex
tmp0 = tl.load(in_ptr0 + (r1 + (16*x0)), xmask, other=0.0)
tmp1 = tl.broadcast_to(tmp0, [XBLOCK, RBLOCK])
tmp3 = tl.where(xmask, tmp1, float("-inf"))
tmp4 = triton_helpers.max2(tmp3, 1)[:, None]
tmp5 = tmp0 - tmp4
tmp6 = tl_math.exp(tmp5)
tmp7 = tl.broadcast_to(tmp6, [XBLOCK, RBLOCK])
tmp9 = tl.where(xmask, tmp7, 0)
tmp10 = tl.sum(tmp9, 1)[:, None]
tmp11 = tmp6 / tmp10
tl.store(out_ptr2 + (r1 + (16*x0)), tmp11, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/3q/c3qwzxbgp3tfdtnw6bsr4tuwglusbyzafblzuo6nppze6njb6ysp.py
# Topologically Sorted Source Nodes: [g_1], Original ATen: [aten.sum]
# Source node to ATen node mapping:
# g_1 => sum_2
# Graph fragment:
# %sum_2 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%view_2, [2]), kwargs = {})
triton_per_fused_sum_2 = async_compile.triton('triton_per_fused_sum_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[16, 16],
reduction_hint=ReductionHint.DEFAULT,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused_sum_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 1, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused_sum_2(in_ptr0, in_ptr1, out_ptr0, xnumel, rnumel, XBLOCK : tl.constexpr):
xnumel = 16
rnumel = 16
RBLOCK: tl.constexpr = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
roffset = 0
rmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
r2 = rindex
x1 = (xindex // 4)
x3 = xindex
tmp0 = tl.load(in_ptr0 + (r2 + (16*x1)), xmask, eviction_policy='evict_last', other=0.0)
tmp1 = tl.load(in_ptr1 + (r2 + (16*x3)), xmask, other=0.0)
tmp2 = tmp0 * tmp1
tmp3 = tl.broadcast_to(tmp2, [XBLOCK, RBLOCK])
tmp5 = tl.where(xmask, tmp3, 0)
tmp6 = tl.sum(tmp5, 1)[:, None]
tl.store(out_ptr0 + (x3), tmp6, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_3, (1, 4, 1, 1), (4, 1, 1, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [add], Original ATen: [aten.add]
stream0 = get_raw_stream(0)
triton_poi_fused_add_0.run(primals_1, primals_2, buf0, 256, grid=grid(256), stream=stream0)
del primals_2
# Topologically Sorted Source Nodes: [c], Original ATen: [aten.convolution]
buf1 = extern_kernels.convolution(buf0, primals_3, stride=(1, 1), padding=(0, 0), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf1, (4, 1, 4, 4), (16, 16, 4, 1))
buf4 = empty_strided_cuda((4, 1, 16), (16, 16, 1), torch.float32)
# Topologically Sorted Source Nodes: [softmax], Original ATen: [aten._softmax]
triton_per_fused__softmax_1.run(buf1, buf4, 4, 16, grid=grid(4), stream=stream0)
del buf1
buf5 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [g_1], Original ATen: [aten.sum]
triton_per_fused_sum_2.run(buf4, primals_1, buf5, 16, 16, grid=grid(16), stream=stream0)
return (reinterpret_tensor(buf4, (4, 1, 4, 4), (16, 16, 4, 1), 0), buf5, primals_1, primals_3, buf0, buf4, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((1, 4, 1, 1), (4, 1, 1, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
class LinearAttention2d(torch.nn.Module):
"""
Linear attention based on parametrized compatibility score function with softmax normalization.
"""
def __init__(self, in_features, out_features):
super(LinearAttention2d, self).__init__()
self.in_features = in_features
self.out_features = out_features
self.proj = None
if self.in_features != self.out_features:
self.proj = torch.nn.Conv2d(self.in_features, self.out_features,
kernel_size=(1, 1), padding=(0, 0), bias=False)
self.score = torch.nn.Conv2d(self.out_features, out_channels=1,
kernel_size=(1, 1), padding=(0, 0), bias=False)
def forward(self, x, g):
b, _, h, w = x.size()
if self.proj is not None:
g = self.proj(g)
c = self.score(x + g)
a = torch.softmax(c.view(b, 1, -1), dim=2).view(b, 1, h, w)
g = torch.mul(a.expand_as(x), x)
g = g.view(b, self.out_features, -1).sum(dim=2)
return a, g
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'in_features': 4, 'out_features': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import math as tl_math
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_add_0(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK: tl.
constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + x0, xmask)
tmp1 = tl.load(in_ptr1 + x0, xmask)
tmp2 = tmp0 + tmp1
tl.store(out_ptr0 + x0, tmp2, xmask)
@triton.jit
def triton_per_fused__softmax_1(in_ptr0, out_ptr2, xnumel, rnumel, XBLOCK:
tl.constexpr):
xnumel = 4
RBLOCK: tl.constexpr = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
r1 = rindex
x0 = xindex
tmp0 = tl.load(in_ptr0 + (r1 + 16 * x0), xmask, other=0.0)
tmp1 = tl.broadcast_to(tmp0, [XBLOCK, RBLOCK])
tmp3 = tl.where(xmask, tmp1, float('-inf'))
tmp4 = triton_helpers.max2(tmp3, 1)[:, None]
tmp5 = tmp0 - tmp4
tmp6 = tl_math.exp(tmp5)
tmp7 = tl.broadcast_to(tmp6, [XBLOCK, RBLOCK])
tmp9 = tl.where(xmask, tmp7, 0)
tmp10 = tl.sum(tmp9, 1)[:, None]
tmp11 = tmp6 / tmp10
tl.store(out_ptr2 + (r1 + 16 * x0), tmp11, xmask)
@triton.jit
def triton_per_fused_sum_2(in_ptr0, in_ptr1, out_ptr0, xnumel, rnumel,
XBLOCK: tl.constexpr):
xnumel = 16
RBLOCK: tl.constexpr = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
r2 = rindex
x1 = xindex // 4
x3 = xindex
tmp0 = tl.load(in_ptr0 + (r2 + 16 * x1), xmask, eviction_policy=
'evict_last', other=0.0)
tmp1 = tl.load(in_ptr1 + (r2 + 16 * x3), xmask, other=0.0)
tmp2 = tmp0 * tmp1
tmp3 = tl.broadcast_to(tmp2, [XBLOCK, RBLOCK])
tmp5 = tl.where(xmask, tmp3, 0)
tmp6 = tl.sum(tmp5, 1)[:, None]
tl.store(out_ptr0 + x3, tmp6, xmask)
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_3, (1, 4, 1, 1), (4, 1, 1, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_add_0[grid(256)](primals_1, primals_2, buf0, 256,
XBLOCK=256, num_warps=4, num_stages=1)
del primals_2
buf1 = extern_kernels.convolution(buf0, primals_3, stride=(1, 1),
padding=(0, 0), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf1, (4, 1, 4, 4), (16, 16, 4, 1))
buf4 = empty_strided_cuda((4, 1, 16), (16, 16, 1), torch.float32)
triton_per_fused__softmax_1[grid(4)](buf1, buf4, 4, 16, XBLOCK=1,
num_warps=2, num_stages=1)
del buf1
buf5 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
triton_per_fused_sum_2[grid(16)](buf4, primals_1, buf5, 16, 16,
XBLOCK=1, num_warps=2, num_stages=1)
return reinterpret_tensor(buf4, (4, 1, 4, 4), (16, 16, 4, 1), 0
), buf5, primals_1, primals_3, buf0, buf4
class LinearAttention2dNew(torch.nn.Module):
"""
Linear attention based on parametrized compatibility score function with softmax normalization.
"""
def __init__(self, in_features, out_features):
super(LinearAttention2dNew, self).__init__()
self.in_features = in_features
self.out_features = out_features
self.proj = None
if self.in_features != self.out_features:
self.proj = torch.nn.Conv2d(self.in_features, self.out_features,
kernel_size=(1, 1), padding=(0, 0), bias=False)
self.score = torch.nn.Conv2d(self.out_features, out_channels=1,
kernel_size=(1, 1), padding=(0, 0), bias=False)
def forward(self, input_0, input_1):
primals_3 = self.score.weight
primals_1 = input_0
primals_2 = input_1
output = call([primals_1, primals_2, primals_3])
return output[0], output[1]
| itsfrank98/CT-COVID | LinearAttention2d | false | 6,911 | [
"MIT"
] | 1 | 3f054000ca0518be2486cf00cfab695b09e39a26 | https://github.com/itsfrank98/CT-COVID/tree/3f054000ca0518be2486cf00cfab695b09e39a26 | import torch
class Model(torch.nn.Module):
"""
Linear attention based on parametrized compatibility score function with softmax normalization.
"""
def __init__(self, in_features, out_features):
super().__init__()
self.in_features = in_features
self.out_features = out_features
self.proj = None
if self.in_features != self.out_features:
self.proj = torch.nn.Conv2d(self.in_features, self.out_features,
kernel_size=(1, 1), padding=(0, 0), bias=False)
self.score = torch.nn.Conv2d(self.out_features, out_channels=1,
kernel_size=(1, 1), padding=(0, 0), bias=False)
def forward(self, x, g):
b, _, h, w = x.size()
if self.proj is not None:
g = self.proj(g)
c = self.score(x + g)
a = torch.softmax(c.view(b, 1, -1), dim=2).view(b, 1, h, w)
g = torch.mul(a.expand_as(x), x)
g = g.view(b, self.out_features, -1).sum(dim=2)
return a, g
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4, 4]
|
UpConv2x2 | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/7v/c7vsd3fryw4ameuaa7ye3zxgcno6raioyn3ogh4j4hjuesnfsqac.py
# Topologically Sorted Source Nodes: [x, x_1], Original ATen: [aten._unsafe_index, aten.constant_pad_nd]
# Source node to ATen node mapping:
# x => _unsafe_index
# x_1 => constant_pad_nd
# Graph fragment:
# %_unsafe_index : [num_users=1] = call_function[target=torch.ops.aten._unsafe_index.Tensor](args = (%primals_1, [None, None, %unsqueeze, %convert_element_type_1]), kwargs = {})
# %constant_pad_nd : [num_users=2] = call_function[target=torch.ops.aten.constant_pad_nd.default](args = (%_unsafe_index, [0, 1, 0, 1], 0.0), kwargs = {})
triton_poi_fused__unsafe_index_constant_pad_nd_0 = async_compile.triton('triton_poi_fused__unsafe_index_constant_pad_nd_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[2048],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__unsafe_index_constant_pad_nd_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 0, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__unsafe_index_constant_pad_nd_0(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 1296
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = (xindex // 9) % 9
x0 = xindex % 9
x2 = (xindex // 81)
x4 = xindex
tmp0 = x1
tmp1 = tl.full([1], 8, tl.int64)
tmp2 = tmp0 < tmp1
tmp3 = x0
tmp4 = tmp3 < tmp1
tmp5 = tmp2 & tmp4
tmp6 = tmp0.to(tl.float32)
tmp7 = 0.5
tmp8 = tmp6 * tmp7
tmp9 = tmp8.to(tl.int32)
tmp10 = tmp3.to(tl.float32)
tmp11 = tmp10 * tmp7
tmp12 = tmp11.to(tl.int32)
tmp13 = tl.load(in_ptr0 + (tmp12 + (4*tmp9) + (16*x2)), tmp5 & xmask, eviction_policy='evict_last', other=0.0)
tl.store(out_ptr0 + (x4), tmp13, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/uf/cuf5jt2t3ewasuxhx2kl62ekxyz4dzt3qli7sc5r22gbgxw37jtb.py
# Topologically Sorted Source Nodes: [x_2], Original ATen: [aten.convolution]
# Source node to ATen node mapping:
# x_2 => convolution
# Graph fragment:
# %convolution : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%constant_pad_nd, %primals_2, %primals_3, [1, 1], [0, 0], [1, 1], False, [0, 0], 1), kwargs = {})
triton_poi_fused_convolution_1 = async_compile.triton('triton_poi_fused_convolution_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[512],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_1', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_1(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 512
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = (xindex // 64) % 2
tmp0 = tl.load(in_out_ptr0 + (x3), xmask)
tmp1 = tl.load(in_ptr0 + (x1), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + (x3), tmp2, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (2, 4, 2, 2), (16, 4, 2, 1))
assert_size_stride(primals_3, (2, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 9, 9), (324, 81, 9, 1), torch.float32)
# Topologically Sorted Source Nodes: [x, x_1], Original ATen: [aten._unsafe_index, aten.constant_pad_nd]
stream0 = get_raw_stream(0)
triton_poi_fused__unsafe_index_constant_pad_nd_0.run(primals_1, buf0, 1296, grid=grid(1296), stream=stream0)
del primals_1
# Topologically Sorted Source Nodes: [x_2], Original ATen: [aten.convolution]
buf1 = extern_kernels.convolution(buf0, primals_2, stride=(1, 1), padding=(0, 0), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf1, (4, 2, 8, 8), (128, 64, 8, 1))
buf2 = buf1; del buf1 # reuse
# Topologically Sorted Source Nodes: [x_2], Original ATen: [aten.convolution]
triton_poi_fused_convolution_1.run(buf2, primals_3, 512, grid=grid(512), stream=stream0)
del primals_3
return (buf2, primals_2, buf0, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((2, 4, 2, 2), (16, 4, 2, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((2, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.utils.data
from torch.cuda import *
class UpConv2x2(nn.Module):
def __init__(self, channels):
super(UpConv2x2, self).__init__()
self.conv = nn.Conv2d(channels, channels // 2, kernel_size=2,
stride=1, padding=0, bias=True)
def forward(self, x):
x = F.interpolate(x, scale_factor=2, mode='nearest')
x = F.pad(x, (0, 1, 0, 1))
x = self.conv(x)
return x
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'channels': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
import torch.nn as nn
import torch.utils.data
from torch.cuda import *
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_poi_fused__unsafe_index_constant_pad_nd_0(in_ptr0, out_ptr0,
xnumel, XBLOCK: tl.constexpr):
xnumel = 1296
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = xindex // 9 % 9
x0 = xindex % 9
x2 = xindex // 81
x4 = xindex
tmp0 = x1
tmp1 = tl.full([1], 8, tl.int64)
tmp2 = tmp0 < tmp1
tmp3 = x0
tmp4 = tmp3 < tmp1
tmp5 = tmp2 & tmp4
tmp6 = tmp0.to(tl.float32)
tmp7 = 0.5
tmp8 = tmp6 * tmp7
tmp9 = tmp8.to(tl.int32)
tmp10 = tmp3.to(tl.float32)
tmp11 = tmp10 * tmp7
tmp12 = tmp11.to(tl.int32)
tmp13 = tl.load(in_ptr0 + (tmp12 + 4 * tmp9 + 16 * x2), tmp5 & xmask,
eviction_policy='evict_last', other=0.0)
tl.store(out_ptr0 + x4, tmp13, xmask)
@triton.jit
def triton_poi_fused_convolution_1(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl
.constexpr):
xnumel = 512
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = xindex // 64 % 2
tmp0 = tl.load(in_out_ptr0 + x3, xmask)
tmp1 = tl.load(in_ptr0 + x1, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + x3, tmp2, xmask)
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (2, 4, 2, 2), (16, 4, 2, 1))
assert_size_stride(primals_3, (2,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 9, 9), (324, 81, 9, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused__unsafe_index_constant_pad_nd_0[grid(1296)](primals_1,
buf0, 1296, XBLOCK=128, num_warps=4, num_stages=1)
del primals_1
buf1 = extern_kernels.convolution(buf0, primals_2, stride=(1, 1),
padding=(0, 0), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf1, (4, 2, 8, 8), (128, 64, 8, 1))
buf2 = buf1
del buf1
triton_poi_fused_convolution_1[grid(512)](buf2, primals_3, 512,
XBLOCK=256, num_warps=4, num_stages=1)
del primals_3
return buf2, primals_2, buf0
class UpConv2x2New(nn.Module):
def __init__(self, channels):
super(UpConv2x2New, self).__init__()
self.conv = nn.Conv2d(channels, channels // 2, kernel_size=2,
stride=1, padding=0, bias=True)
def forward(self, input_0):
primals_2 = self.conv.weight
primals_3 = self.conv.bias
primals_1 = input_0
output = call([primals_1, primals_2, primals_3])
return output[0]
| jabae/detectEM | UpConv2x2 | false | 6,912 | [
"MIT"
] | 1 | 2d1a5116164d0bed0a8ea767a227d05a8970a448 | https://github.com/jabae/detectEM/tree/2d1a5116164d0bed0a8ea767a227d05a8970a448 | import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.utils.data
from torch.cuda import *
class Model(nn.Module):
def __init__(self, channels):
super().__init__()
self.conv = nn.Conv2d(channels, channels // 2, kernel_size=2,
stride=1, padding=0, bias=True)
def forward(self, x):
x = F.interpolate(x, scale_factor=2, mode='nearest')
x = F.pad(x, (0, 1, 0, 1))
x = self.conv(x)
return x
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4]
|
MultiheadAttention | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/dk/cdk4odz276xorciau5ehgl7f3s2mgkf3hrye6xep6kzubczdeqqy.py
# Topologically Sorted Source Nodes: [matmul], Original ATen: [aten.clone]
# Source node to ATen node mapping:
# matmul => clone
# Graph fragment:
# %clone : [num_users=1] = call_function[target=torch.ops.aten.clone.default](args = (%expand,), kwargs = {memory_format: torch.contiguous_format})
triton_poi_fused_clone_0 = async_compile.triton('triton_poi_fused_clone_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16, 4], tile_hint=TileHint.DEFAULT,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clone_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clone_0(in_ptr0, in_ptr1, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = (yindex // 4)
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + (4*x2) + (16*y1)), xmask & ymask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (y0), ymask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(out_ptr0 + (x2 + (4*y3)), tmp2, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/s2/cs2rk3o3kmhydx4oijp6rsdb5atcrq5axy4adadrpl7gkt7scies.py
# Topologically Sorted Source Nodes: [weights], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# weights => exp
# Graph fragment:
# %mul_tensor : [num_users=2] = call_function[target=torch.ops.aten.mul.Tensor](args = (%view_11, 1), kwargs = {})
# %amax_default : [num_users=1] = call_function[target=torch.ops.aten.amax.default](args = (%mul_tensor, [-1], True), kwargs = {})
# %sub_tensor : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%mul_tensor, %amax_default), kwargs = {})
# %mul_tensor_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sub_tensor, 1.0), kwargs = {})
# %exp : [num_users=2] = call_function[target=torch.ops.aten.exp.default](args = (%mul_tensor_1,), kwargs = {})
triton_poi_fused__softmax_1 = async_compile.triton('triton_poi_fused__softmax_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_1(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp3 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp8 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp11 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp1 = 1.0
tmp2 = tmp0 * tmp1
tmp4 = tmp3 * tmp1
tmp6 = tmp5 * tmp1
tmp7 = triton_helpers.maximum(tmp4, tmp6)
tmp9 = tmp8 * tmp1
tmp10 = triton_helpers.maximum(tmp7, tmp9)
tmp12 = tmp11 * tmp1
tmp13 = triton_helpers.maximum(tmp10, tmp12)
tmp14 = tmp2 - tmp13
tmp15 = tmp14 * tmp1
tmp16 = tl_math.exp(tmp15)
tl.store(out_ptr0 + (x2), tmp16, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/3f/c3fx6bzkalkw7u7askqdnz4rzlcoyqiec4r434sjc5x3axxgkrmr.py
# Topologically Sorted Source Nodes: [weights], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# weights => div, sum_1
# Graph fragment:
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%exp, [-1], True), kwargs = {})
# %div : [num_users=2] = call_function[target=torch.ops.aten.div.Tensor](args = (%exp, %sum_1), kwargs = {})
triton_poi_fused__softmax_2 = async_compile.triton('triton_poi_fused__softmax_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_2(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + (x2), tmp8, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/xt/cxtkkmujo4ytg6ycpz5lk5livtstr63pg5nsf5ijewjbtrfrqx6k.py
# Topologically Sorted Source Nodes: [out_1], Original ATen: [aten.clone]
# Source node to ATen node mapping:
# out_1 => clone_3
# Graph fragment:
# %clone_3 : [num_users=1] = call_function[target=torch.ops.aten.clone.default](args = (%view_15,), kwargs = {memory_format: torch.contiguous_format})
triton_poi_fused_clone_3 = async_compile.triton('triton_poi_fused_clone_3', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16, 4], tile_hint=TileHint.SQUARE,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clone_3', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clone_3(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = (yindex // 4)
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + (4*x2) + (16*y1)), xmask & ymask, eviction_policy='evict_last')
tl.store(out_ptr0 + (x2 + (4*y3)), tmp0, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/q4/cq4lrbjfvbivmpg2zkxhkatw7yc2rqarfj625cpqjlxqgfutfyet.py
# Topologically Sorted Source Nodes: [out_1], Original ATen: [aten.add]
# Source node to ATen node mapping:
# out_1 => add
# Graph fragment:
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%view_17, %primals_11), kwargs = {})
triton_poi_fused_add_4 = async_compile.triton('triton_poi_fused_add_4', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_4', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_4(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + (x2), tmp2, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9, primals_10, primals_11 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_2, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_3, (4, 4), (4, 1))
assert_size_stride(primals_4, (4, ), (1, ))
assert_size_stride(primals_5, (4, 4), (4, 1))
assert_size_stride(primals_6, (4, ), (1, ))
assert_size_stride(primals_7, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_8, (4, 4), (4, 1))
assert_size_stride(primals_9, (4, ), (1, ))
assert_size_stride(primals_10, (4, 4), (4, 1))
assert_size_stride(primals_11, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(primals_2, (16, 4), (4, 1), 0), reinterpret_tensor(primals_3, (4, 4), (1, 4), 0), out=buf0)
del primals_3
buf1 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(primals_7, (16, 4), (4, 1), 0), reinterpret_tensor(primals_5, (4, 4), (1, 4), 0), out=buf1)
del primals_5
buf2 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(primals_1, (16, 4), (4, 1), 0), reinterpret_tensor(primals_8, (4, 4), (1, 4), 0), out=buf2)
del primals_8
buf3 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 1), torch.float32)
# Topologically Sorted Source Nodes: [matmul], Original ATen: [aten.clone]
stream0 = get_raw_stream(0)
triton_poi_fused_clone_0.run(buf2, primals_9, buf3, 16, 4, grid=grid(16, 4), stream=stream0)
del primals_9
buf4 = reinterpret_tensor(buf2, (4, 4, 1, 4), (16, 4, 4, 1), 0); del buf2 # reuse
# Topologically Sorted Source Nodes: [matmul], Original ATen: [aten.clone]
triton_poi_fused_clone_0.run(buf0, primals_4, buf4, 16, 4, grid=grid(16, 4), stream=stream0)
del primals_4
buf5 = empty_strided_cuda((16, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [matmul], Original ATen: [aten.bmm]
extern_kernels.bmm(reinterpret_tensor(buf3, (16, 4, 1), (4, 1, 0), 0), reinterpret_tensor(buf4, (16, 1, 4), (4, 0, 1), 0), out=buf5)
buf6 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [weights], Original ATen: [aten._softmax]
triton_poi_fused__softmax_1.run(buf5, buf6, 256, grid=grid(256), stream=stream0)
buf7 = reinterpret_tensor(buf5, (4, 4, 4, 4), (64, 16, 4, 1), 0); del buf5 # reuse
# Topologically Sorted Source Nodes: [weights], Original ATen: [aten._softmax]
triton_poi_fused__softmax_2.run(buf6, buf7, 256, grid=grid(256), stream=stream0)
del buf6
buf8 = reinterpret_tensor(buf0, (4, 4, 4, 1), (16, 4, 1, 1), 0); del buf0 # reuse
# Topologically Sorted Source Nodes: [out], Original ATen: [aten.clone]
triton_poi_fused_clone_0.run(buf1, primals_6, buf8, 16, 4, grid=grid(16, 4), stream=stream0)
del primals_6
buf9 = reinterpret_tensor(buf1, (16, 4, 1), (4, 1, 1), 0); del buf1 # reuse
# Topologically Sorted Source Nodes: [out], Original ATen: [aten.bmm]
extern_kernels.bmm(reinterpret_tensor(buf7, (16, 4, 4), (16, 4, 1), 0), reinterpret_tensor(buf8, (16, 4, 1), (4, 1, 0), 0), out=buf9)
buf10 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [out_1], Original ATen: [aten.clone]
triton_poi_fused_clone_3.run(buf9, buf10, 16, 4, grid=grid(16, 4), stream=stream0)
buf11 = reinterpret_tensor(buf9, (16, 4), (4, 1), 0); del buf9 # reuse
# Topologically Sorted Source Nodes: [out_1], Original ATen: [aten.mm]
extern_kernels.mm(reinterpret_tensor(buf10, (16, 4), (4, 1), 0), reinterpret_tensor(primals_10, (4, 4), (1, 4), 0), out=buf11)
buf12 = reinterpret_tensor(buf11, (4, 4, 4), (16, 4, 1), 0); del buf11 # reuse
# Topologically Sorted Source Nodes: [out_1], Original ATen: [aten.add]
triton_poi_fused_add_4.run(buf12, primals_11, 64, grid=grid(64), stream=stream0)
del primals_11
return (buf12, reinterpret_tensor(primals_2, (16, 4), (4, 1), 0), reinterpret_tensor(primals_7, (16, 4), (4, 1), 0), reinterpret_tensor(primals_1, (16, 4), (4, 1), 0), buf7, reinterpret_tensor(buf10, (16, 4), (4, 1), 0), primals_10, reinterpret_tensor(buf8, (16, 1, 4), (4, 1, 1), 0), reinterpret_tensor(buf3, (16, 1, 4), (4, 1, 1), 0), reinterpret_tensor(buf4, (16, 4, 1), (4, 1, 4), 0), )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_8 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_9 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_10 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_11 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9, primals_10, primals_11])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import numpy as np
from typing import Optional
import torch.nn as nn
class MultiheadAttention(nn.Module):
"""Multihead scaled dot-product attention.
"""
def __init__(self, contexts: 'int', queries: 'int', channels: 'int',
heads: 'int'):
"""Initializer.
Args:
contexts: size of the key, value channels.
queries: size of the query channels.
channels: size of the hidden channels.
heads: the number of the attnetion heads.
"""
super().__init__()
self.channels, self.heads = channels // heads, heads
self.proj_key = nn.Linear(contexts, channels)
self.proj_value = nn.Linear(contexts, channels)
self.proj_query = nn.Linear(queries, channels)
self.proj_out = nn.Linear(channels, channels)
def forward(self, query: 'torch.Tensor', key: 'torch.Tensor', value:
'torch.Tensor', mask: 'Optional[torch.Tensor]'=None) ->torch.Tensor:
"""Transform the inputs.
Args:
query: [torch.float32; [B, S, queries]], query.
key: [torch.float32; [B, T, contexts]], key.
value: [torch.float32; [B, T, contexts]], value.
mask: [torch.float32; [B, S, T]], attention mask.
Returns:
[torch.float32; [B, S, C]], attended.
"""
bsize, querylen, _ = query.shape
keylen = key.shape[1]
key = self.proj_key(key).view(bsize, keylen, self.heads, self.channels)
value = self.proj_value(value).view(bsize, keylen, self.heads, self
.channels)
query = self.proj_query(query).view(bsize, querylen, self.heads,
self.channels)
score = torch.matmul(query.permute(0, 2, 1, 3), key.permute(0, 2, 3, 1)
) * self.channels ** -0.5
if mask is not None:
score.masked_fill_(~mask[:, None, 0:1], -np.inf)
weights = torch.softmax(score, dim=-1)
out = torch.matmul(weights, value.transpose(1, 2))
out = self.proj_out(out.transpose(1, 2).reshape(bsize, querylen, -1))
if mask is not None:
out = out * mask[..., 0:1]
return out
def get_inputs():
return [torch.rand([4, 4, 4]), torch.rand([4, 4, 4]), torch.rand([4, 4, 4])
]
def get_init_inputs():
return [[], {'contexts': 4, 'queries': 4, 'channels': 4, 'heads': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import math as tl_math
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_clone_0(in_ptr0, in_ptr1, out_ptr0, ynumel, xnumel,
YBLOCK: tl.constexpr, XBLOCK: tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = yindex // 4
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + 4 * x2 + 16 * y1), xmask & ymask,
eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + y0, ymask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(out_ptr0 + (x2 + 4 * y3), tmp2, xmask & ymask)
@triton.jit
def triton_poi_fused__softmax_1(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp3 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp8 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp11 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last'
)
tmp1 = 1.0
tmp2 = tmp0 * tmp1
tmp4 = tmp3 * tmp1
tmp6 = tmp5 * tmp1
tmp7 = triton_helpers.maximum(tmp4, tmp6)
tmp9 = tmp8 * tmp1
tmp10 = triton_helpers.maximum(tmp7, tmp9)
tmp12 = tmp11 * tmp1
tmp13 = triton_helpers.maximum(tmp10, tmp12)
tmp14 = tmp2 - tmp13
tmp15 = tmp14 * tmp1
tmp16 = tl_math.exp(tmp15)
tl.store(out_ptr0 + x2, tmp16, xmask)
@triton.jit
def triton_poi_fused__softmax_2(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + x2, tmp8, xmask)
@triton.jit
def triton_poi_fused_clone_3(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK: tl.
constexpr, XBLOCK: tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = yindex // 4
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + 4 * x2 + 16 * y1), xmask & ymask,
eviction_policy='evict_last')
tl.store(out_ptr0 + (x2 + 4 * y3), tmp0, xmask & ymask)
@triton.jit
def triton_poi_fused_add_4(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + x2, tmp2, xmask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7, primals_8, primals_9, primals_10, primals_11) = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_2, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_3, (4, 4), (4, 1))
assert_size_stride(primals_4, (4,), (1,))
assert_size_stride(primals_5, (4, 4), (4, 1))
assert_size_stride(primals_6, (4,), (1,))
assert_size_stride(primals_7, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_8, (4, 4), (4, 1))
assert_size_stride(primals_9, (4,), (1,))
assert_size_stride(primals_10, (4, 4), (4, 1))
assert_size_stride(primals_11, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_2, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_3, (4, 4), (1, 4), 0), out=buf0)
del primals_3
buf1 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_7, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_5, (4, 4), (1, 4), 0), out=buf1)
del primals_5
buf2 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_1, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_8, (4, 4), (1, 4), 0), out=buf2)
del primals_8
buf3 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_clone_0[grid(16, 4)](buf2, primals_9, buf3, 16, 4,
XBLOCK=2, YBLOCK=16, num_warps=1, num_stages=1)
del primals_9
buf4 = reinterpret_tensor(buf2, (4, 4, 1, 4), (16, 4, 4, 1), 0)
del buf2
triton_poi_fused_clone_0[grid(16, 4)](buf0, primals_4, buf4, 16, 4,
XBLOCK=2, YBLOCK=16, num_warps=1, num_stages=1)
del primals_4
buf5 = empty_strided_cuda((16, 4, 4), (16, 4, 1), torch.float32)
extern_kernels.bmm(reinterpret_tensor(buf3, (16, 4, 1), (4, 1, 0),
0), reinterpret_tensor(buf4, (16, 1, 4), (4, 0, 1), 0), out=buf5)
buf6 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
triton_poi_fused__softmax_1[grid(256)](buf5, buf6, 256, XBLOCK=256,
num_warps=4, num_stages=1)
buf7 = reinterpret_tensor(buf5, (4, 4, 4, 4), (64, 16, 4, 1), 0)
del buf5
triton_poi_fused__softmax_2[grid(256)](buf6, buf7, 256, XBLOCK=256,
num_warps=4, num_stages=1)
del buf6
buf8 = reinterpret_tensor(buf0, (4, 4, 4, 1), (16, 4, 1, 1), 0)
del buf0
triton_poi_fused_clone_0[grid(16, 4)](buf1, primals_6, buf8, 16, 4,
XBLOCK=2, YBLOCK=16, num_warps=1, num_stages=1)
del primals_6
buf9 = reinterpret_tensor(buf1, (16, 4, 1), (4, 1, 1), 0)
del buf1
extern_kernels.bmm(reinterpret_tensor(buf7, (16, 4, 4), (16, 4, 1),
0), reinterpret_tensor(buf8, (16, 4, 1), (4, 1, 0), 0), out=buf9)
buf10 = empty_strided_cuda((4, 4, 4), (16, 4, 1), torch.float32)
triton_poi_fused_clone_3[grid(16, 4)](buf9, buf10, 16, 4, XBLOCK=4,
YBLOCK=16, num_warps=1, num_stages=1)
buf11 = reinterpret_tensor(buf9, (16, 4), (4, 1), 0)
del buf9
extern_kernels.mm(reinterpret_tensor(buf10, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_10, (4, 4), (1, 4), 0), out=buf11)
buf12 = reinterpret_tensor(buf11, (4, 4, 4), (16, 4, 1), 0)
del buf11
triton_poi_fused_add_4[grid(64)](buf12, primals_11, 64, XBLOCK=64,
num_warps=1, num_stages=1)
del primals_11
return buf12, reinterpret_tensor(primals_2, (16, 4), (4, 1), 0
), reinterpret_tensor(primals_7, (16, 4), (4, 1), 0
), reinterpret_tensor(primals_1, (16, 4), (4, 1), 0
), buf7, reinterpret_tensor(buf10, (16, 4), (4, 1), 0
), primals_10, reinterpret_tensor(buf8, (16, 1, 4), (4, 1, 1), 0
), reinterpret_tensor(buf3, (16, 1, 4), (4, 1, 1), 0
), reinterpret_tensor(buf4, (16, 4, 1), (4, 1, 4), 0)
class MultiheadAttentionNew(nn.Module):
"""Multihead scaled dot-product attention.
"""
def __init__(self, contexts: 'int', queries: 'int', channels: 'int',
heads: 'int'):
"""Initializer.
Args:
contexts: size of the key, value channels.
queries: size of the query channels.
channels: size of the hidden channels.
heads: the number of the attnetion heads.
"""
super().__init__()
self.channels, self.heads = channels // heads, heads
self.proj_key = nn.Linear(contexts, channels)
self.proj_value = nn.Linear(contexts, channels)
self.proj_query = nn.Linear(queries, channels)
self.proj_out = nn.Linear(channels, channels)
def forward(self, input_0, input_1, input_2):
primals_3 = self.proj_key.weight
primals_4 = self.proj_key.bias
primals_5 = self.proj_value.weight
primals_6 = self.proj_value.bias
primals_8 = self.proj_query.weight
primals_9 = self.proj_query.bias
primals_10 = self.proj_out.weight
primals_11 = self.proj_out.bias
primals_1 = input_0
primals_2 = input_1
primals_7 = input_2
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7, primals_8, primals_9,
primals_10, primals_11])
return output[0]
| ishine/torch-retriever-vc | MultiheadAttention | false | 6,913 | [
"MIT"
] | 1 | db5119d9d703ea819e2ac9185871ea3db52c14e1 | https://github.com/ishine/torch-retriever-vc/tree/db5119d9d703ea819e2ac9185871ea3db52c14e1 | import torch
import numpy as np
from typing import Optional
import torch.nn as nn
class Model(nn.Module):
"""Multihead scaled dot-product attention.
"""
def __init__(self, contexts: 'int', queries: 'int', channels: 'int',
heads: 'int'):
"""Initializer.
Args:
contexts: size of the key, value channels.
queries: size of the query channels.
channels: size of the hidden channels.
heads: the number of the attnetion heads.
"""
super().__init__()
self.channels, self.heads = channels // heads, heads
self.proj_key = nn.Linear(contexts, channels)
self.proj_value = nn.Linear(contexts, channels)
self.proj_query = nn.Linear(queries, channels)
self.proj_out = nn.Linear(channels, channels)
def forward(self, query: 'torch.Tensor', key: 'torch.Tensor', value:
'torch.Tensor', mask: 'Optional[torch.Tensor]'=None) ->torch.Tensor:
"""Transform the inputs.
Args:
query: [torch.float32; [B, S, queries]], query.
key: [torch.float32; [B, T, contexts]], key.
value: [torch.float32; [B, T, contexts]], value.
mask: [torch.float32; [B, S, T]], attention mask.
Returns:
[torch.float32; [B, S, C]], attended.
"""
bsize, querylen, _ = query.shape
keylen = key.shape[1]
key = self.proj_key(key).view(bsize, keylen, self.heads, self.channels)
value = self.proj_value(value).view(bsize, keylen, self.heads, self
.channels)
query = self.proj_query(query).view(bsize, querylen, self.heads,
self.channels)
score = torch.matmul(query.permute(0, 2, 1, 3), key.permute(0, 2, 3, 1)
) * self.channels ** -0.5
if mask is not None:
score.masked_fill_(~mask[:, None, 0:1], -np.inf)
weights = torch.softmax(score, dim=-1)
out = torch.matmul(weights, value.transpose(1, 2))
out = self.proj_out(out.transpose(1, 2).reshape(bsize, querylen, -1))
if mask is not None:
out = out * mask[..., 0:1]
return out
def get_inputs():
return [torch.rand([4, 4, 4]), torch.rand([4, 4, 4]), torch.rand([4, 4, 4])
]
def get_init_inputs():
return [4, 4, 4, 4]
|
Bilinear | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/55/c556jlky4dn3zysohrmkzzgx6ib2mru2umit7utahhvakbnltdhb.py
# Topologically Sorted Source Nodes: [matmul_1], Original ATen: [aten.clone]
# Source node to ATen node mapping:
# matmul_1 => clone
# Graph fragment:
# %clone : [num_users=1] = call_function[target=torch.ops.aten.clone.default](args = (%expand,), kwargs = {memory_format: torch.contiguous_format})
triton_poi_fused_clone_0 = async_compile.triton('triton_poi_fused_clone_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[1024],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clone_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clone_0(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 1024
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 64
x2 = (xindex // 256)
x3 = xindex
tmp0 = tl.load(in_ptr0 + (x0 + (64*x2)), xmask, eviction_policy='evict_last')
tl.store(out_ptr0 + (x3), tmp0, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/c5/cc5niyk56fa3m2oc45eqgnf3eeposiu56eefgo6q2ozqlq3nwxnc.py
# Topologically Sorted Source Nodes: [matmul_1], Original ATen: [aten.clone]
# Source node to ATen node mapping:
# matmul_1 => clone_1
# Graph fragment:
# %clone_1 : [num_users=1] = call_function[target=torch.ops.aten.clone.default](args = (%expand_1,), kwargs = {memory_format: torch.contiguous_format})
triton_poi_fused_clone_1 = async_compile.triton('triton_poi_fused_clone_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[1024],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clone_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clone_1(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 1024
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 16
x2 = (xindex // 64)
x3 = xindex
tmp0 = tl.load(in_ptr0 + (x0 + (16*x2)), xmask, eviction_policy='evict_last')
tl.store(out_ptr0 + (x3), tmp0, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (4, 4), (4, 1))
assert_size_stride(primals_2, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [matmul], Original ATen: [aten.mm]
extern_kernels.mm(reinterpret_tensor(primals_2, (64, 4), (4, 1), 0), primals_1, out=buf0)
del primals_1
buf1 = empty_strided_cuda((4, 4, 4, 4, 4), (256, 64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [matmul_1], Original ATen: [aten.clone]
stream0 = get_raw_stream(0)
triton_poi_fused_clone_0.run(buf0, buf1, 1024, grid=grid(1024), stream=stream0)
del buf0
buf2 = empty_strided_cuda((4, 4, 4, 4, 4), (256, 64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [matmul_1], Original ATen: [aten.clone]
triton_poi_fused_clone_1.run(primals_3, buf2, 1024, grid=grid(1024), stream=stream0)
del primals_3
buf3 = empty_strided_cuda((64, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [matmul_1], Original ATen: [aten.bmm]
extern_kernels.bmm(reinterpret_tensor(buf1, (64, 4, 4), (16, 4, 1), 0), reinterpret_tensor(buf2, (64, 4, 4), (16, 4, 1), 0), out=buf3)
del buf1
return (reinterpret_tensor(buf3, (1024, 1), (1, 1), 0), reinterpret_tensor(buf2, (64, 4, 4), (16, 1, 4), 0), reinterpret_tensor(primals_2, (4, 64), (1, 4), 0), )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
class Bilinear(nn.Module):
def __init__(self, size):
super(Bilinear, self).__init__()
self.size = size
self.mat = nn.Parameter(torch.FloatTensor(self.size, self.size))
self.reset_parameters()
def reset_parameters(self):
params = [p for p in self.parameters() if p.requires_grad]
for i, param in enumerate(params):
param.data.normal_()
def forward(self, vector1, vector2):
bma = torch.matmul(vector1, self.mat).unsqueeze(1)
ba = torch.matmul(bma, vector2.unsqueeze(2)).view(-1, 1)
return ba
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'size': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_clone_0(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 1024
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 64
x2 = xindex // 256
x3 = xindex
tmp0 = tl.load(in_ptr0 + (x0 + 64 * x2), xmask, eviction_policy=
'evict_last')
tl.store(out_ptr0 + x3, tmp0, xmask)
@triton.jit
def triton_poi_fused_clone_1(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 1024
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 16
x2 = xindex // 64
x3 = xindex
tmp0 = tl.load(in_ptr0 + (x0 + 16 * x2), xmask, eviction_policy=
'evict_last')
tl.store(out_ptr0 + x3, tmp0, xmask)
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (4, 4), (4, 1))
assert_size_stride(primals_2, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_2, (64, 4), (4, 1), 0),
primals_1, out=buf0)
del primals_1
buf1 = empty_strided_cuda((4, 4, 4, 4, 4), (256, 64, 16, 4, 1),
torch.float32)
get_raw_stream(0)
triton_poi_fused_clone_0[grid(1024)](buf0, buf1, 1024, XBLOCK=128,
num_warps=4, num_stages=1)
del buf0
buf2 = empty_strided_cuda((4, 4, 4, 4, 4), (256, 64, 16, 4, 1),
torch.float32)
triton_poi_fused_clone_1[grid(1024)](primals_3, buf2, 1024, XBLOCK=
256, num_warps=4, num_stages=1)
del primals_3
buf3 = empty_strided_cuda((64, 4, 4), (16, 4, 1), torch.float32)
extern_kernels.bmm(reinterpret_tensor(buf1, (64, 4, 4), (16, 4, 1),
0), reinterpret_tensor(buf2, (64, 4, 4), (16, 4, 1), 0), out=buf3)
del buf1
return reinterpret_tensor(buf3, (1024, 1), (1, 1), 0), reinterpret_tensor(
buf2, (64, 4, 4), (16, 1, 4), 0), reinterpret_tensor(primals_2, (4,
64), (1, 4), 0)
class BilinearNew(nn.Module):
def __init__(self, size):
super(BilinearNew, self).__init__()
self.size = size
self.mat = nn.Parameter(torch.FloatTensor(self.size, self.size))
self.reset_parameters()
def reset_parameters(self):
params = [p for p in self.parameters() if p.requires_grad]
for i, param in enumerate(params):
param.data.normal_()
def forward(self, input_0, input_1):
primals_1 = self.mat
primals_2 = input_0
primals_3 = input_1
output = call([primals_1, primals_2, primals_3])
return output[0]
| jalshr21/diora-1 | Bilinear | false | 6,914 | [
"Apache-2.0"
] | 1 | a9b680fde6a840707340e9e8232643b0f0e637bd | https://github.com/jalshr21/diora-1/tree/a9b680fde6a840707340e9e8232643b0f0e637bd | import torch
import torch.nn as nn
class Model(nn.Module):
def __init__(self, size):
super().__init__()
self.size = size
self.mat = nn.Parameter(torch.FloatTensor(self.size, self.size))
self.reset_parameters()
def reset_parameters(self):
params = [p for p in self.parameters() if p.requires_grad]
for i, param in enumerate(params):
param.data.normal_()
def forward(self, vector1, vector2):
bma = torch.matmul(vector1, self.mat).unsqueeze(1)
ba = torch.matmul(bma, vector2.unsqueeze(2)).view(-1, 1)
return ba
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4]
|
SigmoidFocalLoss | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/ny/cny2lybv3eo5gjecnhptlj5fm6l572ym36s5acsilpa4ilaczkmy.py
# Topologically Sorted Source Nodes: [pred_sigmoid, sub, pow_1, ne, mask, mul_1, sub_1, pos_part, mul_4, pow_2, neg, max_val, neg_1, exp, neg_2, sub_2, exp_1, add, log, add_1, neg_part, mul_5, add_2, sum_1, neg_3, loss, loss_1], Original ATen: [aten.sigmoid, aten.rsub, aten.pow, aten.ne, aten._to_copy, aten.mul, aten.sub, aten.neg, aten.clamp, aten.exp, aten.add, aten.log, aten.sum, aten.mean]
# Source node to ATen node mapping:
# add => add
# add_1 => add_1
# add_2 => add_2
# exp => exp
# exp_1 => exp_1
# log => log
# loss => mul_6
# loss_1 => mean
# mask => convert_element_type
# max_val => clamp_min
# mul_1 => mul_1
# mul_4 => mul_4
# mul_5 => mul_5
# ne => ne
# neg => neg
# neg_1 => neg_1
# neg_2 => neg_2
# neg_3 => neg_3
# neg_part => mul_3
# pos_part => mul_2
# pow_1 => pow_1
# pow_2 => pow_2
# pred_sigmoid => sigmoid
# sub => sub
# sub_1 => sub_1
# sub_2 => sub_2
# sum_1 => sum_1
# Graph fragment:
# %sigmoid : [num_users=6] = call_function[target=torch.ops.aten.sigmoid.default](args = (%view,), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (1, %sigmoid), kwargs = {})
# %pow_1 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%sub, 2.0), kwargs = {})
# %ne : [num_users=1] = call_function[target=torch.ops.aten.ne.Scalar](args = (%view_1, 4), kwargs = {})
# %convert_element_type : [num_users=2] = call_function[target=torch.ops.prims.convert_element_type.default](args = (%ne, torch.float32), kwargs = {})
# %mul_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sigmoid, %view_2), kwargs = {})
# %sub_1 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%sigmoid, %mul_1), kwargs = {})
# %mul_2 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%pow_1, %sub_1), kwargs = {})
# %mul_4 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%mul_2, 0.25), kwargs = {})
# %pow_2 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%sigmoid, 2.0), kwargs = {})
# %neg : [num_users=1] = call_function[target=torch.ops.aten.neg.default](args = (%sigmoid,), kwargs = {})
# %clamp_min : [num_users=3] = call_function[target=torch.ops.aten.clamp_min.default](args = (%neg, 0), kwargs = {})
# %neg_1 : [num_users=1] = call_function[target=torch.ops.aten.neg.default](args = (%clamp_min,), kwargs = {})
# %exp : [num_users=1] = call_function[target=torch.ops.aten.exp.default](args = (%neg_1,), kwargs = {})
# %neg_2 : [num_users=1] = call_function[target=torch.ops.aten.neg.default](args = (%sigmoid,), kwargs = {})
# %sub_2 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%neg_2, %clamp_min), kwargs = {})
# %exp_1 : [num_users=1] = call_function[target=torch.ops.aten.exp.default](args = (%sub_2,), kwargs = {})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%exp, %exp_1), kwargs = {})
# %log : [num_users=1] = call_function[target=torch.ops.aten.log.default](args = (%add,), kwargs = {})
# %add_1 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%clamp_min, %log), kwargs = {})
# %mul_3 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%pow_2, %add_1), kwargs = {})
# %mul_5 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%mul_3, 0.75), kwargs = {})
# %add_2 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul_4, %mul_5), kwargs = {})
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%add_2, [-1]), kwargs = {})
# %neg_3 : [num_users=1] = call_function[target=torch.ops.aten.neg.default](args = (%sum_1,), kwargs = {})
# %mul_6 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%neg_3, %convert_element_type), kwargs = {})
# %mean : [num_users=1] = call_function[target=torch.ops.aten.mean.default](args = (%mul_6,), kwargs = {})
triton_per_fused__to_copy_add_clamp_exp_log_mean_mul_ne_neg_pow_rsub_sigmoid_sub_sum_0 = async_compile.triton('triton_per_fused__to_copy_add_clamp_exp_log_mean_mul_ne_neg_pow_rsub_sigmoid_sub_sum_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[1, 64],
reduction_hint=ReductionHint.INNER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {3: 1}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 4), equal_to_1=(3,))]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused__to_copy_add_clamp_exp_log_mean_mul_ne_neg_pow_rsub_sigmoid_sub_sum_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 1, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused__to_copy_add_clamp_exp_log_mean_mul_ne_neg_pow_rsub_sigmoid_sub_sum_0(in_out_ptr0, in_ptr0, in_ptr1, xnumel, rnumel, XBLOCK : tl.constexpr):
xnumel = 1
rnumel = 64
RBLOCK: tl.constexpr = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[None, :]
roffset = 0
rmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
r0 = rindex
tmp0 = tl.load(in_ptr0 + (r0), None)
tmp5 = tl.load(in_ptr1 + (r0), None)
tmp1 = tl.sigmoid(tmp0)
tmp2 = 1.0
tmp3 = tmp2 - tmp1
tmp4 = tmp3 * tmp3
tmp6 = 4.0
tmp7 = tmp5 != tmp6
tmp8 = tmp7.to(tl.float32)
tmp9 = tmp8 * tmp5
tmp10 = tmp1 * tmp9
tmp11 = tmp1 - tmp10
tmp12 = tmp4 * tmp11
tmp13 = 0.25
tmp14 = tmp12 * tmp13
tmp15 = tmp1 * tmp1
tmp16 = -tmp1
tmp17 = 0.0
tmp18 = triton_helpers.maximum(tmp16, tmp17)
tmp19 = -tmp18
tmp20 = tl_math.exp(tmp19)
tmp21 = tmp16 - tmp18
tmp22 = tl_math.exp(tmp21)
tmp23 = tmp20 + tmp22
tmp24 = tl_math.log(tmp23)
tmp25 = tmp18 + tmp24
tmp26 = tmp15 * tmp25
tmp27 = 0.75
tmp28 = tmp26 * tmp27
tmp29 = tmp14 + tmp28
tmp30 = -tmp29
tmp31 = tmp30 * tmp8
tmp32 = tl.broadcast_to(tmp31, [XBLOCK, RBLOCK])
tmp34 = tl.sum(tmp32, 1)[:, None]
tmp35 = 64.0
tmp36 = tmp34 / tmp35
tl.debug_barrier()
tl.store(in_out_ptr0 + (tl.full([XBLOCK, 1], 0, tl.int32)), tmp36, None)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, arg1_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4), (16, 4, 1))
assert_size_stride(arg1_1, (4, 16), (16, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf1 = empty_strided_cuda((), (), torch.float32)
buf2 = buf1; del buf1 # reuse
# Topologically Sorted Source Nodes: [pred_sigmoid, sub, pow_1, ne, mask, mul_1, sub_1, pos_part, mul_4, pow_2, neg, max_val, neg_1, exp, neg_2, sub_2, exp_1, add, log, add_1, neg_part, mul_5, add_2, sum_1, neg_3, loss, loss_1], Original ATen: [aten.sigmoid, aten.rsub, aten.pow, aten.ne, aten._to_copy, aten.mul, aten.sub, aten.neg, aten.clamp, aten.exp, aten.add, aten.log, aten.sum, aten.mean]
stream0 = get_raw_stream(0)
triton_per_fused__to_copy_add_clamp_exp_log_mean_mul_ne_neg_pow_rsub_sigmoid_sub_sum_0.run(buf2, arg1_1, arg0_1, 1, 64, grid=grid(1), stream=stream0)
del arg0_1
del arg1_1
return (buf2, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
arg1_1 = rand_strided((4, 16), (16, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([arg0_1, arg1_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.utils
class SigmoidFocalLoss(nn.Module):
def __init__(self, ignore_label, gamma=2.0, alpha=0.25, reduction='mean'):
super(SigmoidFocalLoss, self).__init__()
self.ignore_label = ignore_label
self.gamma = gamma
self.alpha = alpha
self.reduction = reduction
def forward(self, pred, target):
b, _h, _w = target.size()
pred = pred.view(b, -1, 1)
pred_sigmoid = pred.sigmoid()
target = target.view(b, -1).float()
mask = target.ne(self.ignore_label).float()
target = mask * target
onehot = target.view(b, -1, 1)
max_val = (-pred_sigmoid).clamp(min=0)
pos_part = (1 - pred_sigmoid) ** self.gamma * (pred_sigmoid -
pred_sigmoid * onehot)
neg_part = pred_sigmoid ** self.gamma * (max_val + ((-max_val).exp(
) + (-pred_sigmoid - max_val).exp()).log())
loss = -(self.alpha * pos_part + (1 - self.alpha) * neg_part).sum(dim
=-1) * mask
if self.reduction == 'mean':
loss = loss.mean()
return loss
def get_inputs():
return [torch.rand([4, 16]), torch.rand([4, 4, 4])]
def get_init_inputs():
return [[], {'ignore_label': 4}]
| import torch
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import math as tl_math
import torch.nn as nn
import torch.utils
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_per_fused__to_copy_add_clamp_exp_log_mean_mul_ne_neg_pow_rsub_sigmoid_sub_sum_0(
in_out_ptr0, in_ptr0, in_ptr1, xnumel, rnumel, XBLOCK: tl.constexpr):
RBLOCK: tl.constexpr = 64
xoffset = tl.program_id(0) * XBLOCK
xoffset + tl.arange(0, XBLOCK)[:, None]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[None, :]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
r0 = rindex
tmp0 = tl.load(in_ptr0 + r0, None)
tmp5 = tl.load(in_ptr1 + r0, None)
tmp1 = tl.sigmoid(tmp0)
tmp2 = 1.0
tmp3 = tmp2 - tmp1
tmp4 = tmp3 * tmp3
tmp6 = 4.0
tmp7 = tmp5 != tmp6
tmp8 = tmp7.to(tl.float32)
tmp9 = tmp8 * tmp5
tmp10 = tmp1 * tmp9
tmp11 = tmp1 - tmp10
tmp12 = tmp4 * tmp11
tmp13 = 0.25
tmp14 = tmp12 * tmp13
tmp15 = tmp1 * tmp1
tmp16 = -tmp1
tmp17 = 0.0
tmp18 = triton_helpers.maximum(tmp16, tmp17)
tmp19 = -tmp18
tmp20 = tl_math.exp(tmp19)
tmp21 = tmp16 - tmp18
tmp22 = tl_math.exp(tmp21)
tmp23 = tmp20 + tmp22
tmp24 = tl_math.log(tmp23)
tmp25 = tmp18 + tmp24
tmp26 = tmp15 * tmp25
tmp27 = 0.75
tmp28 = tmp26 * tmp27
tmp29 = tmp14 + tmp28
tmp30 = -tmp29
tmp31 = tmp30 * tmp8
tmp32 = tl.broadcast_to(tmp31, [XBLOCK, RBLOCK])
tmp34 = tl.sum(tmp32, 1)[:, None]
tmp35 = 64.0
tmp36 = tmp34 / tmp35
tl.debug_barrier()
tl.store(in_out_ptr0 + tl.full([XBLOCK, 1], 0, tl.int32), tmp36, None)
def call(args):
arg0_1, arg1_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4), (16, 4, 1))
assert_size_stride(arg1_1, (4, 16), (16, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf1 = empty_strided_cuda((), (), torch.float32)
buf2 = buf1
del buf1
get_raw_stream(0)
triton_per_fused__to_copy_add_clamp_exp_log_mean_mul_ne_neg_pow_rsub_sigmoid_sub_sum_0[
grid(1)](buf2, arg1_1, arg0_1, 1, 64, XBLOCK=1, num_warps=2,
num_stages=1)
del arg0_1
del arg1_1
return buf2,
class SigmoidFocalLossNew(nn.Module):
def __init__(self, ignore_label, gamma=2.0, alpha=0.25, reduction='mean'):
super(SigmoidFocalLossNew, self).__init__()
self.ignore_label = ignore_label
self.gamma = gamma
self.alpha = alpha
self.reduction = reduction
def forward(self, input_0, input_1):
arg1_1 = input_0
arg0_1 = input_1
output = call([arg0_1, arg1_1])
return output[0]
| jameslong95/FasterSeg | SigmoidFocalLoss | false | 6,915 | [
"MIT"
] | 1 | 872e04964ea46494a6018d9915cee5476e361c27 | https://github.com/jameslong95/FasterSeg/tree/872e04964ea46494a6018d9915cee5476e361c27 | import torch
import torch.nn as nn
import torch.utils
class Model(nn.Module):
def __init__(self, ignore_label, gamma=2.0, alpha=0.25, reduction='mean'):
super().__init__()
self.ignore_label = ignore_label
self.gamma = gamma
self.alpha = alpha
self.reduction = reduction
def forward(self, pred, target):
b, _h, _w = target.size()
pred = pred.view(b, -1, 1)
pred_sigmoid = pred.sigmoid()
target = target.view(b, -1).float()
mask = target.ne(self.ignore_label).float()
target = mask * target
onehot = target.view(b, -1, 1)
max_val = (-pred_sigmoid).clamp(min=0)
pos_part = (1 - pred_sigmoid) ** self.gamma * (pred_sigmoid -
pred_sigmoid * onehot)
neg_part = pred_sigmoid ** self.gamma * (max_val + ((-max_val).exp(
) + (-pred_sigmoid - max_val).exp()).log())
loss = -(self.alpha * pos_part + (1 - self.alpha) * neg_part).sum(dim
=-1) * mask
if self.reduction == 'mean':
loss = loss.mean()
return loss
def get_inputs():
return [torch.rand([4, 16]), torch.rand([4, 4, 4])]
def get_init_inputs():
return [4]
|
ScaledDotProductAttention | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/s6/cs6dvpj3wfu7iz5hr4tnecveryidweyqejemqvqenqwg3yr2dmlb.py
# Topologically Sorted Source Nodes: [q], Original ATen: [aten.div]
# Source node to ATen node mapping:
# q => div
# Graph fragment:
# %div : [num_users=2] = call_function[target=torch.ops.aten.div.Tensor](args = (%arg0_1, 2.0), kwargs = {})
# %copy_ : [num_users=0] = call_function[target=torch.ops.aten.copy_.default](args = (%arg0_1, %div), kwargs = {})
triton_poi_fused_div_0 = async_compile.triton('triton_poi_fused_div_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_div_0', 'mutated_arg_names': ['in_ptr0', 'out_ptr1'], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_div_0(in_ptr0, out_ptr0, out_ptr1, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (x0), xmask)
tmp1 = 0.5
tmp2 = tmp0 * tmp1
tl.store(out_ptr0 + (x0), tmp2, xmask)
tl.store(out_ptr1 + (x0), tmp2, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/hz/chz2sqsqk26mwhf2dxhgh44jfpu2er5yqjftwkzfav5ctqtx5e7f.py
# Topologically Sorted Source Nodes: [alignment], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# alignment => amax, exp, sub
# Graph fragment:
# %amax : [num_users=1] = call_function[target=torch.ops.aten.amax.default](args = (%view_3, [-1], True), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%view_3, %amax), kwargs = {})
# %exp : [num_users=2] = call_function[target=torch.ops.aten.exp.default](args = (%sub,), kwargs = {})
triton_poi_fused__softmax_1 = async_compile.triton('triton_poi_fused__softmax_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_1(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp3 = triton_helpers.maximum(tmp1, tmp2)
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp7 = triton_helpers.maximum(tmp5, tmp6)
tmp8 = tmp0 - tmp7
tmp9 = tl_math.exp(tmp8)
tl.store(out_ptr0 + (x2), tmp9, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/3f/c3fx6bzkalkw7u7askqdnz4rzlcoyqiec4r434sjc5x3axxgkrmr.py
# Topologically Sorted Source Nodes: [alignment], Original ATen: [aten._softmax]
# Source node to ATen node mapping:
# alignment => div_1, sum_1
# Graph fragment:
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%exp, [-1], True), kwargs = {})
# %div_1 : [num_users=2] = call_function[target=torch.ops.aten.div.Tensor](args = (%exp, %sum_1), kwargs = {})
triton_poi_fused__softmax_2 = async_compile.triton('triton_poi_fused__softmax_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__softmax_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__softmax_2(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + (x2), tmp8, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, arg1_1, arg2_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg1_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg2_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [q], Original ATen: [aten.div]
stream0 = get_raw_stream(0)
triton_poi_fused_div_0.run(arg0_1, buf0, arg0_1, 256, grid=grid(256), stream=stream0)
del arg0_1
buf1 = empty_strided_cuda((16, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [energy], Original ATen: [aten.bmm]
extern_kernels.bmm(reinterpret_tensor(buf0, (16, 4, 4), (16, 4, 1), 0), reinterpret_tensor(arg1_1, (16, 4, 4), (16, 1, 4), 0), out=buf1)
del arg1_1
buf2 = buf0; del buf0 # reuse
# Topologically Sorted Source Nodes: [alignment], Original ATen: [aten._softmax]
triton_poi_fused__softmax_1.run(buf1, buf2, 256, grid=grid(256), stream=stream0)
buf3 = reinterpret_tensor(buf1, (4, 4, 4, 4), (64, 16, 4, 1), 0); del buf1 # reuse
# Topologically Sorted Source Nodes: [alignment], Original ATen: [aten._softmax]
triton_poi_fused__softmax_2.run(buf2, buf3, 256, grid=grid(256), stream=stream0)
buf4 = reinterpret_tensor(buf2, (16, 4, 4), (16, 4, 1), 0); del buf2 # reuse
# Topologically Sorted Source Nodes: [context], Original ATen: [aten.bmm]
extern_kernels.bmm(reinterpret_tensor(buf3, (16, 4, 4), (16, 4, 1), 0), reinterpret_tensor(arg2_1, (16, 4, 4), (16, 4, 1), 0), out=buf4)
del arg2_1
return (reinterpret_tensor(buf4, (4, 4, 4, 4), (64, 16, 4, 1), 0), buf3, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
arg1_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
arg2_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([arg0_1, arg1_1, arg2_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import numpy as np
import torch.nn as nn
class ScaledDotProductAttention(nn.Module):
def __init__(self, dropout):
super().__init__()
self.dropout = nn.Dropout(dropout)
def forward(self, q, k, v, mask=None, rpe_q=None, rpe_v=None):
"""
Args:
q: query (*, query_len, dim)
k: key (*, key_len, dim)
v: value (*, key_len, dim)
mask: (*, query_len, key_len), True will be masked out
rpe_q : (query_len, key_len, dim)
rpe_v : (query_len, key_len, dim)
Returns:
context: (*, query_len, dim)
alignment: (*, query_len, key_len)
"""
dim = q.shape[-1]
q /= dim ** 0.5
energy = q @ k.transpose(-2, -1)
if rpe_q is not None:
energy += torch.einsum('...qd,qkd->...qk', q, rpe_q)
if mask is not None:
energy = energy.masked_fill(mask, np.NINF)
alignment = torch.softmax(energy, dim=-1)
context = self.dropout(alignment) @ v
if rpe_v is not None:
context += torch.einsum('...qk,qkd->...qd', alignment, rpe_v)
return context, alignment
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4]), torch.rand(
[4, 4, 4, 4])]
def get_init_inputs():
return [[], {'dropout': 0.5}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import math as tl_math
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_div_0(in_ptr0, out_ptr0, out_ptr1, xnumel, XBLOCK: tl.
constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + x0, xmask)
tmp1 = 0.5
tmp2 = tmp0 * tmp1
tl.store(out_ptr0 + x0, tmp2, xmask)
tl.store(out_ptr1 + x0, tmp2, xmask)
@triton.jit
def triton_poi_fused__softmax_1(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp3 = triton_helpers.maximum(tmp1, tmp2)
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp7 = triton_helpers.maximum(tmp5, tmp6)
tmp8 = tmp0 - tmp7
tmp9 = tl_math.exp(tmp8)
tl.store(out_ptr0 + x2, tmp9, xmask)
@triton.jit
def triton_poi_fused__softmax_2(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp3 = tmp1 + tmp2
tmp5 = tmp3 + tmp4
tmp7 = tmp5 + tmp6
tmp8 = tmp0 / tmp7
tl.store(out_ptr0 + x2, tmp8, xmask)
def call(args):
arg0_1, arg1_1, arg2_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg1_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg2_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_div_0[grid(256)](arg0_1, buf0, arg0_1, 256, XBLOCK
=128, num_warps=4, num_stages=1)
del arg0_1
buf1 = empty_strided_cuda((16, 4, 4), (16, 4, 1), torch.float32)
extern_kernels.bmm(reinterpret_tensor(buf0, (16, 4, 4), (16, 4, 1),
0), reinterpret_tensor(arg1_1, (16, 4, 4), (16, 1, 4), 0), out=buf1
)
del arg1_1
buf2 = buf0
del buf0
triton_poi_fused__softmax_1[grid(256)](buf1, buf2, 256, XBLOCK=128,
num_warps=4, num_stages=1)
buf3 = reinterpret_tensor(buf1, (4, 4, 4, 4), (64, 16, 4, 1), 0)
del buf1
triton_poi_fused__softmax_2[grid(256)](buf2, buf3, 256, XBLOCK=256,
num_warps=4, num_stages=1)
buf4 = reinterpret_tensor(buf2, (16, 4, 4), (16, 4, 1), 0)
del buf2
extern_kernels.bmm(reinterpret_tensor(buf3, (16, 4, 4), (16, 4, 1),
0), reinterpret_tensor(arg2_1, (16, 4, 4), (16, 4, 1), 0), out=buf4
)
del arg2_1
return reinterpret_tensor(buf4, (4, 4, 4, 4), (64, 16, 4, 1), 0), buf3
class ScaledDotProductAttentionNew(nn.Module):
def __init__(self, dropout):
super().__init__()
self.dropout = nn.Dropout(dropout)
def forward(self, input_0, input_1, input_2):
arg0_1 = input_0
arg1_1 = input_1
arg2_1 = input_2
output = call([arg0_1, arg1_1, arg2_1])
return output[0], output[1]
| jahsylla/stochastic-cslr | ScaledDotProductAttention | false | 6,916 | [
"MIT"
] | 1 | d12d48ebec34183d939917cda2d54f38593dcddb | https://github.com/jahsylla/stochastic-cslr/tree/d12d48ebec34183d939917cda2d54f38593dcddb | import torch
import numpy as np
import torch.nn as nn
class Model(nn.Module):
def __init__(self, dropout):
super().__init__()
self.dropout = nn.Dropout(dropout)
def forward(self, q, k, v, mask=None, rpe_q=None, rpe_v=None):
"""
Args:
q: query (*, query_len, dim)
k: key (*, key_len, dim)
v: value (*, key_len, dim)
mask: (*, query_len, key_len), True will be masked out
rpe_q : (query_len, key_len, dim)
rpe_v : (query_len, key_len, dim)
Returns:
context: (*, query_len, dim)
alignment: (*, query_len, key_len)
"""
dim = q.shape[-1]
q /= dim ** 0.5
energy = q @ k.transpose(-2, -1)
if rpe_q is not None:
energy += torch.einsum('...qd,qkd->...qk', q, rpe_q)
if mask is not None:
energy = energy.masked_fill(mask, np.NINF)
alignment = torch.softmax(energy, dim=-1)
context = self.dropout(alignment) @ v
if rpe_v is not None:
context += torch.einsum('...qk,qkd->...qd', alignment, rpe_v)
return context, alignment
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4]), torch.rand(
[4, 4, 4, 4])]
def get_init_inputs():
return [0.5]
|
JSloss | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/yk/cyksiyx4m63kznvivbdgjezpl54i7bsrlcgwfkzvyziykjbdmdpa.py
# Topologically Sorted Source Nodes: [add, net, kl_div, kl_div_1, add_1, mul], Original ATen: [aten.add, aten.div, aten.xlogy, aten.mul, aten.sub, aten.sum]
# Source node to ATen node mapping:
# add => add
# add_1 => add_1
# kl_div => div_1, eq, full_default, full_default_1, isnan, log, mul, mul_1, sub, sum_1, where, where_1
# kl_div_1 => div_2, eq_1, full_default_2, full_default_3, isnan_1, log_1, mul_2, mul_3, sub_1, sum_2, where_2, where_3
# mul => mul_4
# net => div
# Graph fragment:
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%arg0_1, %arg1_1), kwargs = {})
# %div : [num_users=10] = call_function[target=torch.ops.aten.div.Tensor](args = (%add, 2.0), kwargs = {})
# %isnan : [num_users=1] = call_function[target=torch.ops.aten.isnan.default](args = (%div,), kwargs = {})
# %full_default_1 : [num_users=1] = call_function[target=torch.ops.aten.full.default](args = ([], nan), kwargs = {dtype: torch.float32, layout: torch.strided, device: cuda:0, pin_memory: False})
# %eq : [num_users=1] = call_function[target=torch.ops.aten.eq.Scalar](args = (%div, 0), kwargs = {})
# %full_default : [num_users=1] = call_function[target=torch.ops.aten.full.default](args = ([], 0.0), kwargs = {dtype: torch.float32, layout: torch.strided, device: cuda:0, pin_memory: False})
# %log : [num_users=1] = call_function[target=torch.ops.aten.log.default](args = (%div,), kwargs = {})
# %mul_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%div, %log), kwargs = {})
# %where : [num_users=1] = call_function[target=torch.ops.aten.where.self](args = (%eq, %full_default, %mul_1), kwargs = {})
# %where_1 : [num_users=1] = call_function[target=torch.ops.aten.where.self](args = (%isnan, %full_default_1, %where), kwargs = {})
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%div, %arg0_1), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%where_1, %mul), kwargs = {})
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.default](args = (%sub,), kwargs = {})
# %div_1 : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%sum_1, 4), kwargs = {})
# %isnan_1 : [num_users=1] = call_function[target=torch.ops.aten.isnan.default](args = (%div,), kwargs = {})
# %full_default_3 : [num_users=1] = call_function[target=torch.ops.aten.full.default](args = ([], nan), kwargs = {dtype: torch.float32, layout: torch.strided, device: cuda:0, pin_memory: False})
# %eq_1 : [num_users=1] = call_function[target=torch.ops.aten.eq.Scalar](args = (%div, 0), kwargs = {})
# %full_default_2 : [num_users=1] = call_function[target=torch.ops.aten.full.default](args = ([], 0.0), kwargs = {dtype: torch.float32, layout: torch.strided, device: cuda:0, pin_memory: False})
# %log_1 : [num_users=1] = call_function[target=torch.ops.aten.log.default](args = (%div,), kwargs = {})
# %mul_3 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%div, %log_1), kwargs = {})
# %where_2 : [num_users=1] = call_function[target=torch.ops.aten.where.self](args = (%eq_1, %full_default_2, %mul_3), kwargs = {})
# %where_3 : [num_users=1] = call_function[target=torch.ops.aten.where.self](args = (%isnan_1, %full_default_3, %where_2), kwargs = {})
# %mul_2 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%div, %arg1_1), kwargs = {})
# %sub_1 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%where_3, %mul_2), kwargs = {})
# %sum_2 : [num_users=1] = call_function[target=torch.ops.aten.sum.default](args = (%sub_1,), kwargs = {})
# %div_2 : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%sum_2, 4), kwargs = {})
# %add_1 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%div_1, %div_2), kwargs = {})
# %mul_4 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%add_1, 0.5), kwargs = {})
triton_per_fused_add_div_mul_sub_sum_xlogy_0 = async_compile.triton('triton_per_fused_add_div_mul_sub_sum_xlogy_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[1, 256],
reduction_hint=ReductionHint.INNER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {3: 1}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 4), equal_to_1=(3,))]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused_add_div_mul_sub_sum_xlogy_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': True, 'num_load': 2, 'num_reduction': 2, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused_add_div_mul_sub_sum_xlogy_0(in_out_ptr0, in_ptr0, in_ptr1, xnumel, rnumel):
xnumel = 1
XBLOCK: tl.constexpr = 1
rnumel = 256
RBLOCK: tl.constexpr = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = tl.full([1], xoffset, tl.int32)
xmask = tl.full([RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[:]
roffset = 0
rmask = tl.full([RBLOCK], True, tl.int1)
r0 = rindex
tmp0 = tl.load(in_ptr0 + (r0), None)
tmp1 = tl.load(in_ptr1 + (r0), None)
tmp2 = tmp0 + tmp1
tmp3 = 0.5
tmp4 = tmp2 * tmp3
tmp5 = libdevice.isnan(tmp4).to(tl.int1)
tmp6 = 0.0
tmp7 = tmp4 == tmp6
tmp8 = tl_math.log(tmp4)
tmp9 = tmp4 * tmp8
tmp10 = tl.where(tmp7, tmp6, tmp9)
tmp11 = float("nan")
tmp12 = tl.where(tmp5, tmp11, tmp10)
tmp13 = tmp4 * tmp0
tmp14 = tmp12 - tmp13
tmp15 = tl.broadcast_to(tmp14, [RBLOCK])
tmp17 = triton_helpers.promote_to_tensor(tl.sum(tmp15, 0))
tmp18 = tmp4 * tmp1
tmp19 = tmp12 - tmp18
tmp20 = tl.broadcast_to(tmp19, [RBLOCK])
tmp22 = triton_helpers.promote_to_tensor(tl.sum(tmp20, 0))
tmp23 = 0.25
tmp24 = tmp17 * tmp23
tmp25 = tmp22 * tmp23
tmp26 = tmp24 + tmp25
tmp27 = tmp26 * tmp3
tl.debug_barrier()
tl.store(in_out_ptr0 + (tl.full([1], 0, tl.int32)), tmp27, None)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, arg1_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg1_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((), (), torch.float32)
buf2 = buf0; del buf0 # reuse
# Topologically Sorted Source Nodes: [add, net, kl_div, kl_div_1, add_1, mul], Original ATen: [aten.add, aten.div, aten.xlogy, aten.mul, aten.sub, aten.sum]
stream0 = get_raw_stream(0)
triton_per_fused_add_div_mul_sub_sum_xlogy_0.run(buf2, arg0_1, arg1_1, 1, 256, grid=grid(1), stream=stream0)
del arg0_1
del arg1_1
return (buf2, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
arg1_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([arg0_1, arg1_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.nn.functional as F
class JSloss(nn.Module):
""" Compute the Jensen-Shannon loss using the torch native kl_div"""
def __init__(self, reduction='batchmean'):
super().__init__()
self.red = reduction
def forward(self, input, target):
net = (input + target) / 2.0
return 0.5 * (F.kl_div(input, net, reduction=self.red) + F.kl_div(
target, net, reduction=self.red))
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {}]
| import torch
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_per_fused_add_div_mul_sub_sum_xlogy_0(in_out_ptr0, in_ptr0,
in_ptr1, xnumel, rnumel):
XBLOCK: tl.constexpr = 1
RBLOCK: tl.constexpr = 256
xoffset = tl.program_id(0) * XBLOCK
tl.full([1], xoffset, tl.int32)
tl.full([RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[:]
tl.full([RBLOCK], True, tl.int1)
r0 = rindex
tmp0 = tl.load(in_ptr0 + r0, None)
tmp1 = tl.load(in_ptr1 + r0, None)
tmp2 = tmp0 + tmp1
tmp3 = 0.5
tmp4 = tmp2 * tmp3
tmp5 = libdevice.isnan(tmp4).to(tl.int1)
tmp6 = 0.0
tmp7 = tmp4 == tmp6
tmp8 = tl_math.log(tmp4)
tmp9 = tmp4 * tmp8
tmp10 = tl.where(tmp7, tmp6, tmp9)
tmp11 = float('nan')
tmp12 = tl.where(tmp5, tmp11, tmp10)
tmp13 = tmp4 * tmp0
tmp14 = tmp12 - tmp13
tmp15 = tl.broadcast_to(tmp14, [RBLOCK])
tmp17 = triton_helpers.promote_to_tensor(tl.sum(tmp15, 0))
tmp18 = tmp4 * tmp1
tmp19 = tmp12 - tmp18
tmp20 = tl.broadcast_to(tmp19, [RBLOCK])
tmp22 = triton_helpers.promote_to_tensor(tl.sum(tmp20, 0))
tmp23 = 0.25
tmp24 = tmp17 * tmp23
tmp25 = tmp22 * tmp23
tmp26 = tmp24 + tmp25
tmp27 = tmp26 * tmp3
tl.debug_barrier()
tl.store(in_out_ptr0 + tl.full([1], 0, tl.int32), tmp27, None)
def call(args):
arg0_1, arg1_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg1_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((), (), torch.float32)
buf2 = buf0
del buf0
get_raw_stream(0)
triton_per_fused_add_div_mul_sub_sum_xlogy_0[grid(1)](buf2, arg0_1,
arg1_1, 1, 256, num_warps=2, num_stages=1)
del arg0_1
del arg1_1
return buf2,
class JSlossNew(nn.Module):
""" Compute the Jensen-Shannon loss using the torch native kl_div"""
def __init__(self, reduction='batchmean'):
super().__init__()
self.red = reduction
def forward(self, input_0, input_1):
arg0_1 = input_0
arg1_1 = input_1
output = call([arg0_1, arg1_1])
return output[0]
| jaredaevans/UltrafastNST | JSloss | false | 6,917 | [
"MIT"
] | 1 | 6671c6b618ce6bb4920b15f782be962e484a5423 | https://github.com/jaredaevans/UltrafastNST/tree/6671c6b618ce6bb4920b15f782be962e484a5423 | import torch
import torch.nn as nn
import torch.nn.functional as F
class Model(nn.Module):
""" Compute the Jensen-Shannon loss using the torch native kl_div"""
def __init__(self, reduction='batchmean'):
super().__init__()
self.red = reduction
def forward(self, input, target):
net = (input + target) / 2.0
return 0.5 * (F.kl_div(input, net, reduction=self.red) + F.kl_div(
target, net, reduction=self.red))
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return []
|
USConv2d | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/sr/csrhhqsexdcor6gq6tz4dawxblhadgekinzxxkt33uwojltligp6.py
# Topologically Sorted Source Nodes: [y], Original ATen: [aten.convolution]
# Source node to ATen node mapping:
# y => convolution
# Graph fragment:
# %convolution : [num_users=1] = call_function[target=torch.ops.aten.convolution.default](args = (%primals_3, %slice_2, %slice_5, [1, 1], [0, 0], [1, 1], False, [0, 0], 1), kwargs = {})
triton_poi_fused_convolution_0 = async_compile.triton('triton_poi_fused_convolution_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_0(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + (x2), tmp2, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (4, ), (1, ))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
# Topologically Sorted Source Nodes: [y], Original ATen: [aten.convolution]
buf0 = extern_kernels.convolution(primals_3, primals_1, stride=(1, 1), padding=(0, 0), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf0, (4, 4, 1, 1), (4, 1, 1, 1))
buf1 = buf0; del buf0 # reuse
# Topologically Sorted Source Nodes: [y], Original ATen: [aten.convolution]
stream0 = get_raw_stream(0)
triton_poi_fused_convolution_0.run(buf1, primals_2, 16, grid=grid(16), stream=stream0)
del primals_2
return (buf1, primals_3, primals_1, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.utils
def make_divisible(v, divisor=8, min_value=1):
"""
forked from slim:
https://github.com/tensorflow/models/blob/ 0344c5503ee55e24f0de7f37336a6e08f10976fd/ research/slim/nets/mobilenet/mobilenet.py#L62-L69
"""
if min_value is None:
min_value = divisor
new_v = max(min_value, int(v + divisor / 2) // divisor * divisor)
if new_v < 0.9 * v:
new_v += divisor
return new_v
class USConv2d(nn.Conv2d):
def __init__(self, in_channels, out_channels, kernel_size, stride=1,
padding=0, dilation=1, groups=1, depthwise=False, bias=True,
width_mult_list=[1.0]):
super(USConv2d, self).__init__(in_channels, out_channels,
kernel_size, stride=stride, padding=padding, dilation=dilation,
groups=groups, bias=bias)
self.depthwise = depthwise
self.in_channels_max = in_channels
self.out_channels_max = out_channels
self.width_mult_list = width_mult_list
self.ratio = 1.0, 1.0
def set_ratio(self, ratio):
self.ratio = ratio
def forward(self, input):
assert self.ratio[0] in self.width_mult_list, str(self.ratio[0]
) + ' in? ' + str(self.width_mult_list)
self.in_channels = make_divisible(self.in_channels_max * self.ratio[0])
assert self.ratio[1] in self.width_mult_list, str(self.ratio[1]
) + ' in? ' + str(self.width_mult_list)
self.out_channels = make_divisible(self.out_channels_max * self.
ratio[1])
self.groups = self.in_channels if self.depthwise else 1
weight = self.weight[:self.out_channels, :self.in_channels, :, :]
if self.bias is not None:
bias = self.bias[:self.out_channels]
else:
bias = self.bias
y = nn.functional.conv2d(input, weight, bias, self.stride, self.
padding, self.dilation, self.groups)
return y
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'in_channels': 4, 'out_channels': 4, 'kernel_size': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
import torch.nn as nn
import torch.utils
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
@triton.jit
def triton_poi_fused_convolution_0(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl
.constexpr):
xnumel = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 4
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + x2, tmp2, xmask)
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (4,), (1,))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = extern_kernels.convolution(primals_3, primals_1, stride=(1,
1), padding=(0, 0), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf0, (4, 4, 1, 1), (4, 1, 1, 1))
buf1 = buf0
del buf0
get_raw_stream(0)
triton_poi_fused_convolution_0[grid(16)](buf1, primals_2, 16,
XBLOCK=16, num_warps=1, num_stages=1)
del primals_2
return buf1, primals_3, primals_1
def make_divisible(v, divisor=8, min_value=1):
"""
forked from slim:
https://github.com/tensorflow/models/blob/ 0344c5503ee55e24f0de7f37336a6e08f10976fd/ research/slim/nets/mobilenet/mobilenet.py#L62-L69
"""
if min_value is None:
min_value = divisor
new_v = max(min_value, int(v + divisor / 2) // divisor * divisor)
if new_v < 0.9 * v:
new_v += divisor
return new_v
class USConv2dNew(nn.Conv2d):
def __init__(self, in_channels, out_channels, kernel_size, stride=1,
padding=0, dilation=1, groups=1, depthwise=False, bias=True,
width_mult_list=[1.0]):
super(USConv2dNew, self).__init__(in_channels, out_channels,
kernel_size, stride=stride, padding=padding, dilation=dilation,
groups=groups, bias=bias)
self.depthwise = depthwise
self.in_channels_max = in_channels
self.out_channels_max = out_channels
self.width_mult_list = width_mult_list
self.ratio = 1.0, 1.0
def set_ratio(self, ratio):
self.ratio = ratio
def forward(self, input_0):
primals_1 = self.weight
primals_2 = self.bias
primals_3 = input_0
output = call([primals_1, primals_2, primals_3])
return output[0]
| jameslong95/FasterSeg | USConv2d | false | 6,918 | [
"MIT"
] | 1 | 872e04964ea46494a6018d9915cee5476e361c27 | https://github.com/jameslong95/FasterSeg/tree/872e04964ea46494a6018d9915cee5476e361c27 | import torch
import torch.nn as nn
import torch.utils
def make_divisible(v, divisor=8, min_value=1):
"""
forked from slim:
https://github.com/tensorflow/models/blob/ 0344c5503ee55e24f0de7f37336a6e08f10976fd/ research/slim/nets/mobilenet/mobilenet.py#L62-L69
"""
if min_value is None:
min_value = divisor
new_v = max(min_value, int(v + divisor / 2) // divisor * divisor)
if new_v < 0.9 * v:
new_v += divisor
return new_v
class Model(nn.Conv2d):
def __init__(self, in_channels, out_channels, kernel_size, stride=1,
padding=0, dilation=1, groups=1, depthwise=False, bias=True,
width_mult_list=[1.0]):
super().__init__(in_channels, out_channels,
kernel_size, stride=stride, padding=padding, dilation=dilation,
groups=groups, bias=bias)
self.depthwise = depthwise
self.in_channels_max = in_channels
self.out_channels_max = out_channels
self.width_mult_list = width_mult_list
self.ratio = 1.0, 1.0
def set_ratio(self, ratio):
self.ratio = ratio
def forward(self, input):
assert self.ratio[0] in self.width_mult_list, str(self.ratio[0]
) + ' in? ' + str(self.width_mult_list)
self.in_channels = make_divisible(self.in_channels_max * self.ratio[0])
assert self.ratio[1] in self.width_mult_list, str(self.ratio[1]
) + ' in? ' + str(self.width_mult_list)
self.out_channels = make_divisible(self.out_channels_max * self.
ratio[1])
self.groups = self.in_channels if self.depthwise else 1
weight = self.weight[:self.out_channels, :self.in_channels, :, :]
if self.bias is not None:
bias = self.bias[:self.out_channels]
else:
bias = self.bias
y = nn.functional.conv2d(input, weight, bias, self.stride, self.
padding, self.dilation, self.groups)
return y
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4, 4, 4]
|
GetStyleLoss | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/7j/c7jkqpq7g244sdbwoxdaraaqjhcray4b7wkphrkwkxlezb6or6wc.py
# Topologically Sorted Source Nodes: [gram_8], Original ATen: [aten.stack]
# Source node to ATen node mapping:
# gram_8 => cat
# Graph fragment:
# %cat : [num_users=1] = call_function[target=torch.ops.aten.cat.default](args = ([%div, %div_1, %div_2, %div_3],), kwargs = {})
triton_poi_fused_stack_0 = async_compile.triton('triton_poi_fused_stack_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: '*fp32', 5: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_stack_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 4, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_stack_0(in_ptr0, in_ptr1, in_ptr2, in_ptr3, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = (xindex // 4)
x0 = xindex % 4
x2 = xindex
tmp0 = x1
tmp1 = tl.full([1], 0, tl.int64)
tmp2 = tmp0 >= tmp1
tmp3 = tl.full([1], 4, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = tl.load(in_ptr0 + (x0 + (4*x1)), tmp4 & xmask, other=0.0)
tmp6 = 0.0625
tmp7 = tmp5 * tmp6
tmp8 = tl.full(tmp7.shape, 0.0, tmp7.dtype)
tmp9 = tl.where(tmp4, tmp7, tmp8)
tmp10 = tmp0 >= tmp3
tmp11 = tl.full([1], 8, tl.int64)
tmp12 = tmp0 < tmp11
tmp13 = tmp10 & tmp12
tmp14 = tl.load(in_ptr1 + (x0 + (4*((-4) + x1))), tmp13 & xmask, other=0.0)
tmp15 = tmp14 * tmp6
tmp16 = tl.full(tmp15.shape, 0.0, tmp15.dtype)
tmp17 = tl.where(tmp13, tmp15, tmp16)
tmp18 = tmp0 >= tmp11
tmp19 = tl.full([1], 12, tl.int64)
tmp20 = tmp0 < tmp19
tmp21 = tmp18 & tmp20
tmp22 = tl.load(in_ptr2 + (x0 + (4*((-8) + x1))), tmp21 & xmask, other=0.0)
tmp23 = tmp22 * tmp6
tmp24 = tl.full(tmp23.shape, 0.0, tmp23.dtype)
tmp25 = tl.where(tmp21, tmp23, tmp24)
tmp26 = tmp0 >= tmp19
tmp27 = tl.full([1], 16, tl.int64)
tmp28 = tmp0 < tmp27
tmp29 = tl.load(in_ptr3 + (x0 + (4*((-12) + x1))), tmp26 & xmask, other=0.0)
tmp30 = tmp29 * tmp6
tmp31 = tl.full(tmp30.shape, 0.0, tmp30.dtype)
tmp32 = tl.where(tmp26, tmp30, tmp31)
tmp33 = tl.where(tmp21, tmp25, tmp32)
tmp34 = tl.where(tmp13, tmp17, tmp33)
tmp35 = tl.where(tmp4, tmp9, tmp34)
tl.store(out_ptr0 + (x2), tmp35, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/nb/cnbsgcp6nb2y4sgoxl77wsqlkfegfab222q4hyzu7iytqqlbb46p.py
# Topologically Sorted Source Nodes: [mse_loss], Original ATen: [aten.mse_loss]
# Source node to ATen node mapping:
# mse_loss => mean, pow_1, sub
# Graph fragment:
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%expand, %arg1_1), kwargs = {})
# %pow_1 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%sub, 2), kwargs = {})
# %mean : [num_users=1] = call_function[target=torch.ops.aten.mean.default](args = (%pow_1,), kwargs = {})
triton_per_fused_mse_loss_1 = async_compile.triton('triton_per_fused_mse_loss_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[1, 256],
reduction_hint=ReductionHint.INNER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {3: 1}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 4), equal_to_1=(3,))]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused_mse_loss_1', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': True, 'num_load': 2, 'num_reduction': 1, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused_mse_loss_1(in_out_ptr0, in_ptr0, in_ptr1, xnumel, rnumel):
xnumel = 1
XBLOCK: tl.constexpr = 1
rnumel = 256
RBLOCK: tl.constexpr = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = tl.full([1], xoffset, tl.int32)
xmask = tl.full([RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[:]
roffset = 0
rmask = tl.full([RBLOCK], True, tl.int1)
r0 = rindex % 64
r2 = rindex
tmp0 = tl.load(in_ptr0 + (r0), None, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (r2), None)
tmp2 = tmp0 - tmp1
tmp3 = tmp2 * tmp2
tmp4 = tl.broadcast_to(tmp3, [RBLOCK])
tmp6 = triton_helpers.promote_to_tensor(tl.sum(tmp4, 0))
tmp7 = 256.0
tmp8 = tmp6 / tmp7
tl.debug_barrier()
tl.store(in_out_ptr0 + (tl.full([1], 0, tl.int32)), tmp8, None)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, arg1_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg1_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [gram], Original ATen: [aten.mm]
extern_kernels.mm(reinterpret_tensor(arg0_1, (4, 16), (16, 1), 0), reinterpret_tensor(arg0_1, (16, 4), (1, 16), 0), out=buf0)
buf1 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [gram_2], Original ATen: [aten.mm]
extern_kernels.mm(reinterpret_tensor(arg0_1, (4, 16), (16, 1), 64), reinterpret_tensor(arg0_1, (16, 4), (1, 16), 64), out=buf1)
buf2 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [gram_4], Original ATen: [aten.mm]
extern_kernels.mm(reinterpret_tensor(arg0_1, (4, 16), (16, 1), 128), reinterpret_tensor(arg0_1, (16, 4), (1, 16), 128), out=buf2)
buf3 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [gram_6], Original ATen: [aten.mm]
extern_kernels.mm(reinterpret_tensor(arg0_1, (4, 16), (16, 1), 192), reinterpret_tensor(arg0_1, (16, 4), (1, 16), 192), out=buf3)
del arg0_1
buf4 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [gram_8], Original ATen: [aten.stack]
stream0 = get_raw_stream(0)
triton_poi_fused_stack_0.run(buf0, buf1, buf2, buf3, buf4, 64, grid=grid(64), stream=stream0)
del buf0
del buf1
del buf2
del buf3
buf5 = empty_strided_cuda((), (), torch.float32)
buf6 = buf5; del buf5 # reuse
# Topologically Sorted Source Nodes: [mse_loss], Original ATen: [aten.mse_loss]
triton_per_fused_mse_loss_1.run(buf6, buf4, arg1_1, 1, 256, grid=grid(1), stream=stream0)
del arg1_1
del buf4
return (buf6, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
arg1_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([arg0_1, arg1_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.nn.functional as F
def gram_matrix(input):
""" gram matrix for feature assignments """
a, b, c, d = input.size()
allG = []
for i in range(a):
features = input[i].view(b, c * d)
gram = torch.mm(features, features.t())
gram = gram.div(c * d)
allG.append(gram)
return torch.stack(allG)
class GetStyleLoss(nn.Module):
""" evaluate the style loss with gram matrix """
def forward(self, input, target):
""" forward pass """
gram = gram_matrix(target)
return F.mse_loss(gram, input)
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_stack_0(in_ptr0, in_ptr1, in_ptr2, in_ptr3, out_ptr0,
xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = xindex // 4
x0 = xindex % 4
x2 = xindex
tmp0 = x1
tl.full([1], 0, tl.int64)
tmp3 = tl.full([1], 4, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = tl.load(in_ptr0 + (x0 + 4 * x1), tmp4 & xmask, other=0.0)
tmp6 = 0.0625
tmp7 = tmp5 * tmp6
tmp8 = tl.full(tmp7.shape, 0.0, tmp7.dtype)
tmp9 = tl.where(tmp4, tmp7, tmp8)
tmp10 = tmp0 >= tmp3
tmp11 = tl.full([1], 8, tl.int64)
tmp12 = tmp0 < tmp11
tmp13 = tmp10 & tmp12
tmp14 = tl.load(in_ptr1 + (x0 + 4 * (-4 + x1)), tmp13 & xmask, other=0.0)
tmp15 = tmp14 * tmp6
tmp16 = tl.full(tmp15.shape, 0.0, tmp15.dtype)
tmp17 = tl.where(tmp13, tmp15, tmp16)
tmp18 = tmp0 >= tmp11
tmp19 = tl.full([1], 12, tl.int64)
tmp20 = tmp0 < tmp19
tmp21 = tmp18 & tmp20
tmp22 = tl.load(in_ptr2 + (x0 + 4 * (-8 + x1)), tmp21 & xmask, other=0.0)
tmp23 = tmp22 * tmp6
tmp24 = tl.full(tmp23.shape, 0.0, tmp23.dtype)
tmp25 = tl.where(tmp21, tmp23, tmp24)
tmp26 = tmp0 >= tmp19
tl.full([1], 16, tl.int64)
tmp29 = tl.load(in_ptr3 + (x0 + 4 * (-12 + x1)), tmp26 & xmask, other=0.0)
tmp30 = tmp29 * tmp6
tmp31 = tl.full(tmp30.shape, 0.0, tmp30.dtype)
tmp32 = tl.where(tmp26, tmp30, tmp31)
tmp33 = tl.where(tmp21, tmp25, tmp32)
tmp34 = tl.where(tmp13, tmp17, tmp33)
tmp35 = tl.where(tmp4, tmp9, tmp34)
tl.store(out_ptr0 + x2, tmp35, xmask)
@triton.jit
def triton_per_fused_mse_loss_1(in_out_ptr0, in_ptr0, in_ptr1, xnumel, rnumel):
XBLOCK: tl.constexpr = 1
RBLOCK: tl.constexpr = 256
xoffset = tl.program_id(0) * XBLOCK
tl.full([1], xoffset, tl.int32)
tl.full([RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[:]
tl.full([RBLOCK], True, tl.int1)
r0 = rindex % 64
r2 = rindex
tmp0 = tl.load(in_ptr0 + r0, None, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + r2, None)
tmp2 = tmp0 - tmp1
tmp3 = tmp2 * tmp2
tmp4 = tl.broadcast_to(tmp3, [RBLOCK])
tmp6 = triton_helpers.promote_to_tensor(tl.sum(tmp4, 0))
tmp7 = 256.0
tmp8 = tmp6 / tmp7
tl.debug_barrier()
tl.store(in_out_ptr0 + tl.full([1], 0, tl.int32), tmp8, None)
def call(args):
arg0_1, arg1_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg1_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(arg0_1, (4, 16), (16, 1), 0),
reinterpret_tensor(arg0_1, (16, 4), (1, 16), 0), out=buf0)
buf1 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(arg0_1, (4, 16), (16, 1), 64),
reinterpret_tensor(arg0_1, (16, 4), (1, 16), 64), out=buf1)
buf2 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(arg0_1, (4, 16), (16, 1), 128),
reinterpret_tensor(arg0_1, (16, 4), (1, 16), 128), out=buf2)
buf3 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(arg0_1, (4, 16), (16, 1), 192),
reinterpret_tensor(arg0_1, (16, 4), (1, 16), 192), out=buf3)
del arg0_1
buf4 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_stack_0[grid(64)](buf0, buf1, buf2, buf3, buf4, 64,
XBLOCK=64, num_warps=1, num_stages=1)
del buf0
del buf1
del buf2
del buf3
buf5 = empty_strided_cuda((), (), torch.float32)
buf6 = buf5
del buf5
triton_per_fused_mse_loss_1[grid(1)](buf6, buf4, arg1_1, 1, 256,
num_warps=2, num_stages=1)
del arg1_1
del buf4
return buf6,
def gram_matrix(input):
""" gram matrix for feature assignments """
a, b, c, d = input.size()
allG = []
for i in range(a):
features = input[i].view(b, c * d)
gram = torch.mm(features, features.t())
gram = gram.div(c * d)
allG.append(gram)
return torch.stack(allG)
class GetStyleLossNew(nn.Module):
""" evaluate the style loss with gram matrix """
def forward(self, input_0, input_1):
arg0_1 = input_0
arg1_1 = input_1
output = call([arg0_1, arg1_1])
return output[0]
| jaredaevans/UltrafastNST | GetStyleLoss | false | 6,919 | [
"MIT"
] | 1 | 6671c6b618ce6bb4920b15f782be962e484a5423 | https://github.com/jaredaevans/UltrafastNST/tree/6671c6b618ce6bb4920b15f782be962e484a5423 | import torch
import torch.nn as nn
import torch.nn.functional as F
def gram_matrix(input):
""" gram matrix for feature assignments """
a, b, c, d = input.size()
allG = []
for i in range(a):
features = input[i].view(b, c * d)
gram = torch.mm(features, features.t())
gram = gram.div(c * d)
allG.append(gram)
return torch.stack(allG)
class Model(nn.Module):
""" evaluate the style loss with gram matrix """
def forward(self, input, target):
""" forward pass """
gram = gram_matrix(target)
return F.mse_loss(gram, input)
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return []
|
ConvBlock | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/pw/cpw5jgywzg5ntkknxkt5orxsrrr5zq7a6eoteboi3ba7zrcxj2p7.py
# Topologically Sorted Source Nodes: [conv2d], Original ATen: [aten.convolution]
# Source node to ATen node mapping:
# conv2d => convolution
# Graph fragment:
# %convolution : [num_users=3] = call_function[target=torch.ops.aten.convolution.default](args = (%primals_3, %primals_1, %primals_2, [1, 1], [1, 1], [1, 1], False, [0, 0], 1), kwargs = {})
triton_poi_fused_convolution_0 = async_compile.triton('triton_poi_fused_convolution_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_0(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = (xindex // 16) % 4
tmp0 = tl.load(in_out_ptr0 + (x3), xmask)
tmp1 = tl.load(in_ptr0 + (x1), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + (x3), tmp2, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/rr/crrw4mva4sqcfnybxs3vkgipqao2sh7wylecqnd2ucobnkl4v7nn.py
# Topologically Sorted Source Nodes: [batch_norm], Original ATen: [aten._native_batch_norm_legit]
# Source node to ATen node mapping:
# batch_norm => add, rsqrt, var_mean
# Graph fragment:
# %var_mean : [num_users=2] = call_function[target=torch.ops.aten.var_mean.correction](args = (%convolution, [0, 2, 3]), kwargs = {correction: 0, keepdim: True})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%getitem, 1e-05), kwargs = {})
# %rsqrt : [num_users=2] = call_function[target=torch.ops.aten.rsqrt.default](args = (%add,), kwargs = {})
triton_per_fused__native_batch_norm_legit_1 = async_compile.triton('triton_per_fused__native_batch_norm_legit_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[4, 64],
reduction_hint=ReductionHint.INNER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32', 5: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 5), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused__native_batch_norm_legit_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 4, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused__native_batch_norm_legit_1(in_ptr0, out_ptr0, out_ptr1, out_ptr2, xnumel, rnumel, XBLOCK : tl.constexpr):
xnumel = 4
rnumel = 64
RBLOCK: tl.constexpr = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
roffset = 0
rmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
r1 = rindex % 16
r2 = (rindex // 16)
x0 = xindex
tmp0 = tl.load(in_ptr0 + (r1 + (16*x0) + (64*r2)), xmask, other=0.0)
tmp1 = tl.broadcast_to(tmp0, [XBLOCK, RBLOCK])
tmp3 = tl.where(xmask, tmp1, 0)
tmp4 = tl.broadcast_to(tmp1, [XBLOCK, RBLOCK])
tmp6 = tl.where(xmask, tmp4, 0)
tmp7 = tl.sum(tmp6, 1)[:, None]
tmp8 = tl.full([XBLOCK, 1], 64, tl.int32)
tmp9 = tmp8.to(tl.float32)
tmp10 = tmp7 / tmp9
tmp11 = tmp1 - tmp10
tmp12 = tmp11 * tmp11
tmp13 = tl.broadcast_to(tmp12, [XBLOCK, RBLOCK])
tmp15 = tl.where(xmask, tmp13, 0)
tmp16 = tl.sum(tmp15, 1)[:, None]
tmp17 = 64.0
tmp18 = tmp16 / tmp17
tmp19 = 1e-05
tmp20 = tmp18 + tmp19
tmp21 = libdevice.rsqrt(tmp20)
tl.store(out_ptr2 + (x0), tmp21, xmask)
tl.store(out_ptr0 + (x0), tmp10, xmask)
tl.store(out_ptr1 + (x0), tmp16, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/qz/cqzgzfujtioszc3jhyh5gc2jqbhryymqfpk6fwywltfi7trspyd5.py
# Topologically Sorted Source Nodes: [batch_norm, x], Original ATen: [aten._native_batch_norm_legit, aten.relu]
# Source node to ATen node mapping:
# batch_norm => add, add_1, mul, mul_1, rsqrt, sub, var_mean
# x => relu
# Graph fragment:
# %var_mean : [num_users=2] = call_function[target=torch.ops.aten.var_mean.correction](args = (%convolution, [0, 2, 3]), kwargs = {correction: 0, keepdim: True})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%getitem, 1e-05), kwargs = {})
# %rsqrt : [num_users=2] = call_function[target=torch.ops.aten.rsqrt.default](args = (%add,), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%convolution, %getitem_1), kwargs = {})
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sub, %rsqrt), kwargs = {})
# %mul_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%mul, %unsqueeze_1), kwargs = {})
# %add_1 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul_1, %unsqueeze_3), kwargs = {})
# %relu : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%add_1,), kwargs = {})
triton_poi_fused__native_batch_norm_legit_relu_2 = async_compile.triton('triton_poi_fused__native_batch_norm_legit_relu_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: '*fp32', 5: '*fp32', 6: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5, 6), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__native_batch_norm_legit_relu_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__native_batch_norm_legit_relu_2(in_ptr0, in_ptr1, in_ptr2, in_ptr3, in_ptr4, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = (xindex // 16) % 4
tmp0 = tl.load(in_ptr0 + (x3), xmask)
tmp1 = tl.load(in_ptr1 + (x1), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr2 + (x1), xmask, eviction_policy='evict_last')
tmp10 = tl.load(in_ptr3 + (x1), xmask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr4 + (x1), xmask, eviction_policy='evict_last')
tmp2 = tmp0 - tmp1
tmp4 = 64.0
tmp5 = tmp3 / tmp4
tmp6 = 1e-05
tmp7 = tmp5 + tmp6
tmp8 = libdevice.rsqrt(tmp7)
tmp9 = tmp2 * tmp8
tmp11 = tmp9 * tmp10
tmp13 = tmp11 + tmp12
tmp14 = tl.full([1], 0, tl.int32)
tmp15 = triton_helpers.maximum(tmp14, tmp13)
tl.store(out_ptr0 + (x3), tmp15, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/vt/cvti3mmkydkye4nbcxadixl24xpri53wfh2hb7signebfdn3bzkl.py
# Topologically Sorted Source Nodes: [batch_norm_2, x_2], Original ATen: [aten._native_batch_norm_legit, aten.relu, aten.threshold_backward]
# Source node to ATen node mapping:
# batch_norm_2 => add_4, add_5, mul_4, mul_5, rsqrt_2, sub_2, var_mean_2
# x_2 => relu_2
# Graph fragment:
# %var_mean_2 : [num_users=2] = call_function[target=torch.ops.aten.var_mean.correction](args = (%convolution_2, [0, 2, 3]), kwargs = {correction: 0, keepdim: True})
# %add_4 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%getitem_4, 1e-05), kwargs = {})
# %rsqrt_2 : [num_users=2] = call_function[target=torch.ops.aten.rsqrt.default](args = (%add_4,), kwargs = {})
# %sub_2 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%convolution_2, %getitem_5), kwargs = {})
# %mul_4 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sub_2, %rsqrt_2), kwargs = {})
# %mul_5 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%mul_4, %unsqueeze_1), kwargs = {})
# %add_5 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul_5, %unsqueeze_3), kwargs = {})
# %relu_2 : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%add_5,), kwargs = {})
# %le : [num_users=1] = call_function[target=torch.ops.aten.le.Scalar](args = (%relu_2, 0), kwargs = {})
triton_poi_fused__native_batch_norm_legit_relu_threshold_backward_3 = async_compile.triton('triton_poi_fused__native_batch_norm_legit_relu_threshold_backward_3', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: '*fp32', 5: '*fp32', 6: '*i1', 7: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5, 6, 7), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__native_batch_norm_legit_relu_threshold_backward_3', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__native_batch_norm_legit_relu_threshold_backward_3(in_ptr0, in_ptr1, in_ptr2, in_ptr3, in_ptr4, out_ptr0, out_ptr1, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = (xindex // 16) % 4
tmp0 = tl.load(in_ptr0 + (x3), xmask)
tmp1 = tl.load(in_ptr1 + (x1), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr2 + (x1), xmask, eviction_policy='evict_last')
tmp10 = tl.load(in_ptr3 + (x1), xmask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr4 + (x1), xmask, eviction_policy='evict_last')
tmp2 = tmp0 - tmp1
tmp4 = 64.0
tmp5 = tmp3 / tmp4
tmp6 = 1e-05
tmp7 = tmp5 + tmp6
tmp8 = libdevice.rsqrt(tmp7)
tmp9 = tmp2 * tmp8
tmp11 = tmp9 * tmp10
tmp13 = tmp11 + tmp12
tmp14 = tl.full([1], 0, tl.int32)
tmp15 = triton_helpers.maximum(tmp14, tmp13)
tmp16 = 0.0
tmp17 = tmp15 <= tmp16
tl.store(out_ptr0 + (x3), tmp15, xmask)
tl.store(out_ptr1 + (x3), tmp17, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 3, 3), (36, 9, 3, 1))
assert_size_stride(primals_2, (4, ), (1, ))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_4, (4, ), (1, ))
assert_size_stride(primals_5, (4, ), (1, ))
assert_size_stride(primals_6, (4, 4, 3, 3), (36, 9, 3, 1))
assert_size_stride(primals_7, (4, ), (1, ))
assert_size_stride(primals_8, (4, 4, 3, 3), (36, 9, 3, 1))
assert_size_stride(primals_9, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
# Topologically Sorted Source Nodes: [conv2d], Original ATen: [aten.convolution]
buf0 = extern_kernels.convolution(primals_3, primals_1, stride=(1, 1), padding=(1, 1), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf0, (4, 4, 4, 4), (64, 16, 4, 1))
buf1 = buf0; del buf0 # reuse
# Topologically Sorted Source Nodes: [conv2d], Original ATen: [aten.convolution]
stream0 = get_raw_stream(0)
triton_poi_fused_convolution_0.run(buf1, primals_2, 256, grid=grid(256), stream=stream0)
del primals_2
buf2 = empty_strided_cuda((1, 4, 1, 1), (4, 1, 4, 4), torch.float32)
buf3 = empty_strided_cuda((1, 4, 1, 1), (4, 1, 4, 4), torch.float32)
buf5 = empty_strided_cuda((1, 4, 1, 1), (4, 1, 4, 4), torch.float32)
# Topologically Sorted Source Nodes: [batch_norm], Original ATen: [aten._native_batch_norm_legit]
triton_per_fused__native_batch_norm_legit_1.run(buf1, buf2, buf3, buf5, 4, 64, grid=grid(4), stream=stream0)
buf6 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [batch_norm, x], Original ATen: [aten._native_batch_norm_legit, aten.relu]
triton_poi_fused__native_batch_norm_legit_relu_2.run(buf1, buf2, buf3, primals_4, primals_5, buf6, 256, grid=grid(256), stream=stream0)
# Topologically Sorted Source Nodes: [conv2d_1], Original ATen: [aten.convolution]
buf7 = extern_kernels.convolution(buf6, primals_6, stride=(1, 1), padding=(1, 1), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf7, (4, 4, 4, 4), (64, 16, 4, 1))
buf8 = buf7; del buf7 # reuse
# Topologically Sorted Source Nodes: [conv2d_1], Original ATen: [aten.convolution]
triton_poi_fused_convolution_0.run(buf8, primals_7, 256, grid=grid(256), stream=stream0)
del primals_7
buf9 = buf3; del buf3 # reuse
buf10 = empty_strided_cuda((1, 4, 1, 1), (4, 1, 4, 4), torch.float32)
buf12 = empty_strided_cuda((1, 4, 1, 1), (4, 1, 4, 4), torch.float32)
# Topologically Sorted Source Nodes: [batch_norm_1], Original ATen: [aten._native_batch_norm_legit]
triton_per_fused__native_batch_norm_legit_1.run(buf8, buf9, buf10, buf12, 4, 64, grid=grid(4), stream=stream0)
buf13 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [batch_norm_1, x_1], Original ATen: [aten._native_batch_norm_legit, aten.relu]
triton_poi_fused__native_batch_norm_legit_relu_2.run(buf8, buf9, buf10, primals_4, primals_5, buf13, 256, grid=grid(256), stream=stream0)
# Topologically Sorted Source Nodes: [conv2d_2], Original ATen: [aten.convolution]
buf14 = extern_kernels.convolution(buf13, primals_8, stride=(1, 1), padding=(1, 1), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf14, (4, 4, 4, 4), (64, 16, 4, 1))
buf15 = buf14; del buf14 # reuse
# Topologically Sorted Source Nodes: [conv2d_2], Original ATen: [aten.convolution]
triton_poi_fused_convolution_0.run(buf15, primals_9, 256, grid=grid(256), stream=stream0)
del primals_9
buf16 = buf10; del buf10 # reuse
buf17 = empty_strided_cuda((1, 4, 1, 1), (4, 1, 4, 4), torch.float32)
buf19 = empty_strided_cuda((1, 4, 1, 1), (4, 1, 4, 4), torch.float32)
# Topologically Sorted Source Nodes: [batch_norm_2], Original ATen: [aten._native_batch_norm_legit]
triton_per_fused__native_batch_norm_legit_1.run(buf15, buf16, buf17, buf19, 4, 64, grid=grid(4), stream=stream0)
buf20 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
buf21 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.bool)
# Topologically Sorted Source Nodes: [batch_norm_2, x_2], Original ATen: [aten._native_batch_norm_legit, aten.relu, aten.threshold_backward]
triton_poi_fused__native_batch_norm_legit_relu_threshold_backward_3.run(buf15, buf16, buf17, primals_4, primals_5, buf20, buf21, 256, grid=grid(256), stream=stream0)
del buf17
del primals_5
return (buf20, primals_1, primals_3, primals_4, primals_6, primals_8, buf1, reinterpret_tensor(buf5, (4, ), (1, ), 0), buf6, buf8, reinterpret_tensor(buf12, (4, ), (1, ), 0), buf13, buf15, reinterpret_tensor(buf19, (4, ), (1, ), 0), buf21, reinterpret_tensor(buf16, (1, 4, 1, 1), (4, 1, 1, 1), 0), reinterpret_tensor(buf9, (1, 4, 1, 1), (4, 1, 1, 1), 0), reinterpret_tensor(buf2, (1, 4, 1, 1), (4, 1, 1, 1), 0), )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 3, 3), (36, 9, 3, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((4, 4, 3, 3), (36, 9, 3, 1), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_8 = rand_strided((4, 4, 3, 3), (36, 9, 3, 1), device='cuda:0', dtype=torch.float32)
primals_9 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.utils.data
from torch.cuda import *
def conv3x3(in_channels, out_channels):
return nn.Conv2d(in_channels, out_channels, kernel_size=3, stride=1,
padding=1, bias=True)
class ConvBlock(nn.Module):
def __init__(self, in_channels, out_channels):
"""
Args:
in_channels: number of channels in input (1st) feature map
out_channels: number of channels in output feature maps
"""
super(ConvBlock, self).__init__()
self.conv1 = conv3x3(in_channels, out_channels)
self.conv2 = conv3x3(out_channels, out_channels)
self.conv3 = conv3x3(out_channels, out_channels)
self.norm = nn.BatchNorm2d(out_channels, track_running_stats=False)
def forward(self, x):
x = F.relu(self.norm(self.conv1(x)))
x = F.relu(self.norm(self.conv2(x)))
x = F.relu(self.norm(self.conv3(x)))
return x
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'in_channels': 4, 'out_channels': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import libdevice
import torch.nn as nn
import torch.utils.data
from torch.cuda import *
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_convolution_0(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl
.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = xindex // 16 % 4
tmp0 = tl.load(in_out_ptr0 + x3, xmask)
tmp1 = tl.load(in_ptr0 + x1, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + x3, tmp2, xmask)
@triton.jit
def triton_per_fused__native_batch_norm_legit_1(in_ptr0, out_ptr0, out_ptr1,
out_ptr2, xnumel, rnumel, XBLOCK: tl.constexpr):
xnumel = 4
RBLOCK: tl.constexpr = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
r1 = rindex % 16
r2 = rindex // 16
x0 = xindex
tmp0 = tl.load(in_ptr0 + (r1 + 16 * x0 + 64 * r2), xmask, other=0.0)
tmp1 = tl.broadcast_to(tmp0, [XBLOCK, RBLOCK])
tl.where(xmask, tmp1, 0)
tmp4 = tl.broadcast_to(tmp1, [XBLOCK, RBLOCK])
tmp6 = tl.where(xmask, tmp4, 0)
tmp7 = tl.sum(tmp6, 1)[:, None]
tmp8 = tl.full([XBLOCK, 1], 64, tl.int32)
tmp9 = tmp8.to(tl.float32)
tmp10 = tmp7 / tmp9
tmp11 = tmp1 - tmp10
tmp12 = tmp11 * tmp11
tmp13 = tl.broadcast_to(tmp12, [XBLOCK, RBLOCK])
tmp15 = tl.where(xmask, tmp13, 0)
tmp16 = tl.sum(tmp15, 1)[:, None]
tmp17 = 64.0
tmp18 = tmp16 / tmp17
tmp19 = 1e-05
tmp20 = tmp18 + tmp19
tmp21 = libdevice.rsqrt(tmp20)
tl.store(out_ptr2 + x0, tmp21, xmask)
tl.store(out_ptr0 + x0, tmp10, xmask)
tl.store(out_ptr1 + x0, tmp16, xmask)
@triton.jit
def triton_poi_fused__native_batch_norm_legit_relu_2(in_ptr0, in_ptr1,
in_ptr2, in_ptr3, in_ptr4, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = xindex // 16 % 4
tmp0 = tl.load(in_ptr0 + x3, xmask)
tmp1 = tl.load(in_ptr1 + x1, xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr2 + x1, xmask, eviction_policy='evict_last')
tmp10 = tl.load(in_ptr3 + x1, xmask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr4 + x1, xmask, eviction_policy='evict_last')
tmp2 = tmp0 - tmp1
tmp4 = 64.0
tmp5 = tmp3 / tmp4
tmp6 = 1e-05
tmp7 = tmp5 + tmp6
tmp8 = libdevice.rsqrt(tmp7)
tmp9 = tmp2 * tmp8
tmp11 = tmp9 * tmp10
tmp13 = tmp11 + tmp12
tmp14 = tl.full([1], 0, tl.int32)
tmp15 = triton_helpers.maximum(tmp14, tmp13)
tl.store(out_ptr0 + x3, tmp15, xmask)
@triton.jit
def triton_poi_fused__native_batch_norm_legit_relu_threshold_backward_3(in_ptr0
, in_ptr1, in_ptr2, in_ptr3, in_ptr4, out_ptr0, out_ptr1, xnumel,
XBLOCK: tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = xindex // 16 % 4
tmp0 = tl.load(in_ptr0 + x3, xmask)
tmp1 = tl.load(in_ptr1 + x1, xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr2 + x1, xmask, eviction_policy='evict_last')
tmp10 = tl.load(in_ptr3 + x1, xmask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr4 + x1, xmask, eviction_policy='evict_last')
tmp2 = tmp0 - tmp1
tmp4 = 64.0
tmp5 = tmp3 / tmp4
tmp6 = 1e-05
tmp7 = tmp5 + tmp6
tmp8 = libdevice.rsqrt(tmp7)
tmp9 = tmp2 * tmp8
tmp11 = tmp9 * tmp10
tmp13 = tmp11 + tmp12
tmp14 = tl.full([1], 0, tl.int32)
tmp15 = triton_helpers.maximum(tmp14, tmp13)
tmp16 = 0.0
tmp17 = tmp15 <= tmp16
tl.store(out_ptr0 + x3, tmp15, xmask)
tl.store(out_ptr1 + x3, tmp17, xmask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7, primals_8, primals_9) = args
args.clear()
assert_size_stride(primals_1, (4, 4, 3, 3), (36, 9, 3, 1))
assert_size_stride(primals_2, (4,), (1,))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_4, (4,), (1,))
assert_size_stride(primals_5, (4,), (1,))
assert_size_stride(primals_6, (4, 4, 3, 3), (36, 9, 3, 1))
assert_size_stride(primals_7, (4,), (1,))
assert_size_stride(primals_8, (4, 4, 3, 3), (36, 9, 3, 1))
assert_size_stride(primals_9, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = extern_kernels.convolution(primals_3, primals_1, stride=(1,
1), padding=(1, 1), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf0, (4, 4, 4, 4), (64, 16, 4, 1))
buf1 = buf0
del buf0
get_raw_stream(0)
triton_poi_fused_convolution_0[grid(256)](buf1, primals_2, 256,
XBLOCK=256, num_warps=4, num_stages=1)
del primals_2
buf2 = empty_strided_cuda((1, 4, 1, 1), (4, 1, 4, 4), torch.float32)
buf3 = empty_strided_cuda((1, 4, 1, 1), (4, 1, 4, 4), torch.float32)
buf5 = empty_strided_cuda((1, 4, 1, 1), (4, 1, 4, 4), torch.float32)
triton_per_fused__native_batch_norm_legit_1[grid(4)](buf1, buf2,
buf3, buf5, 4, 64, XBLOCK=1, num_warps=2, num_stages=1)
buf6 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
triton_poi_fused__native_batch_norm_legit_relu_2[grid(256)](buf1,
buf2, buf3, primals_4, primals_5, buf6, 256, XBLOCK=128,
num_warps=4, num_stages=1)
buf7 = extern_kernels.convolution(buf6, primals_6, stride=(1, 1),
padding=(1, 1), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf7, (4, 4, 4, 4), (64, 16, 4, 1))
buf8 = buf7
del buf7
triton_poi_fused_convolution_0[grid(256)](buf8, primals_7, 256,
XBLOCK=256, num_warps=4, num_stages=1)
del primals_7
buf9 = buf3
del buf3
buf10 = empty_strided_cuda((1, 4, 1, 1), (4, 1, 4, 4), torch.float32)
buf12 = empty_strided_cuda((1, 4, 1, 1), (4, 1, 4, 4), torch.float32)
triton_per_fused__native_batch_norm_legit_1[grid(4)](buf8, buf9,
buf10, buf12, 4, 64, XBLOCK=1, num_warps=2, num_stages=1)
buf13 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
triton_poi_fused__native_batch_norm_legit_relu_2[grid(256)](buf8,
buf9, buf10, primals_4, primals_5, buf13, 256, XBLOCK=128,
num_warps=4, num_stages=1)
buf14 = extern_kernels.convolution(buf13, primals_8, stride=(1, 1),
padding=(1, 1), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf14, (4, 4, 4, 4), (64, 16, 4, 1))
buf15 = buf14
del buf14
triton_poi_fused_convolution_0[grid(256)](buf15, primals_9, 256,
XBLOCK=256, num_warps=4, num_stages=1)
del primals_9
buf16 = buf10
del buf10
buf17 = empty_strided_cuda((1, 4, 1, 1), (4, 1, 4, 4), torch.float32)
buf19 = empty_strided_cuda((1, 4, 1, 1), (4, 1, 4, 4), torch.float32)
triton_per_fused__native_batch_norm_legit_1[grid(4)](buf15, buf16,
buf17, buf19, 4, 64, XBLOCK=1, num_warps=2, num_stages=1)
buf20 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
buf21 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.bool)
triton_poi_fused__native_batch_norm_legit_relu_threshold_backward_3[
grid(256)](buf15, buf16, buf17, primals_4, primals_5, buf20,
buf21, 256, XBLOCK=128, num_warps=4, num_stages=1)
del buf17
del primals_5
return (buf20, primals_1, primals_3, primals_4, primals_6, primals_8,
buf1, reinterpret_tensor(buf5, (4,), (1,), 0), buf6, buf8,
reinterpret_tensor(buf12, (4,), (1,), 0), buf13, buf15,
reinterpret_tensor(buf19, (4,), (1,), 0), buf21, reinterpret_tensor
(buf16, (1, 4, 1, 1), (4, 1, 1, 1), 0), reinterpret_tensor(buf9, (1,
4, 1, 1), (4, 1, 1, 1), 0), reinterpret_tensor(buf2, (1, 4, 1, 1),
(4, 1, 1, 1), 0))
def conv3x3(in_channels, out_channels):
return nn.Conv2d(in_channels, out_channels, kernel_size=3, stride=1,
padding=1, bias=True)
class ConvBlockNew(nn.Module):
def __init__(self, in_channels, out_channels):
"""
Args:
in_channels: number of channels in input (1st) feature map
out_channels: number of channels in output feature maps
"""
super(ConvBlockNew, self).__init__()
self.conv1 = conv3x3(in_channels, out_channels)
self.conv2 = conv3x3(out_channels, out_channels)
self.conv3 = conv3x3(out_channels, out_channels)
self.norm = nn.BatchNorm2d(out_channels, track_running_stats=False)
def forward(self, input_0):
primals_1 = self.conv1.weight
primals_2 = self.conv1.bias
primals_6 = self.conv2.weight
primals_4 = self.conv2.bias
primals_8 = self.conv3.weight
primals_5 = self.conv3.bias
primals_7 = self.norm.weight
primals_9 = self.norm.bias
primals_3 = input_0
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7, primals_8, primals_9])
return output[0]
| jabae/detectEM | ConvBlock | false | 6,920 | [
"MIT"
] | 1 | 2d1a5116164d0bed0a8ea767a227d05a8970a448 | https://github.com/jabae/detectEM/tree/2d1a5116164d0bed0a8ea767a227d05a8970a448 | import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.utils.data
from torch.cuda import *
def conv3x3(in_channels, out_channels):
return nn.Conv2d(in_channels, out_channels, kernel_size=3, stride=1,
padding=1, bias=True)
class Model(nn.Module):
def __init__(self, in_channels, out_channels):
"""
Args:
in_channels: number of channels in input (1st) feature map
out_channels: number of channels in output feature maps
"""
super().__init__()
self.conv1 = conv3x3(in_channels, out_channels)
self.conv2 = conv3x3(out_channels, out_channels)
self.conv3 = conv3x3(out_channels, out_channels)
self.norm = nn.BatchNorm2d(out_channels, track_running_stats=False)
def forward(self, x):
x = F.relu(self.norm(self.conv1(x)))
x = F.relu(self.norm(self.conv2(x)))
x = F.relu(self.norm(self.conv3(x)))
return x
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4, 4]
|
BertAttention | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/x2/cx2hdvwyo7m5jvhhvtugzxqvmy6z4nsfhkkjhvgzbbm3cb6dsum2.py
# Topologically Sorted Source Nodes: [], Original ATen: []
# Source node to ATen node mapping:
# Graph fragment:
# %mul_scalar : [num_users=1] = call_function[target=torch.ops.aten.mul.Scalar](args = (%permute_default, 1.0), kwargs = {})
# %clone_default : [num_users=1] = call_function[target=torch.ops.aten.clone.default](args = (%expand_default,), kwargs = {memory_format: torch.contiguous_format})
triton_poi_fused_0 = async_compile.triton('triton_poi_fused_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16, 4], tile_hint=TileHint.DEFAULT,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_0(in_ptr0, in_ptr1, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = (yindex // 4)
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + (4*x2) + (16*y1)), xmask & ymask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (y0), ymask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = 1.0
tmp4 = tmp2 * tmp3
tl.store(out_ptr0 + (x2 + (4*y3)), tmp4, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/5j/c5jll3kxtd32cl7pwubrb5oky2mtzckfgip2xbwad7crvvp4zk4r.py
# Topologically Sorted Source Nodes: [], Original ATen: []
# Source node to ATen node mapping:
# Graph fragment:
# %amax_default : [num_users=1] = call_function[target=torch.ops.aten.amax.default](args = (%view_default_2, [-1], True), kwargs = {})
# %sub_tensor : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%view_default_2, %amax_default), kwargs = {})
# %exp_default : [num_users=2] = call_function[target=torch.ops.aten.exp.default](args = (%sub_tensor,), kwargs = {})
triton_poi_fused_1 = async_compile.triton('triton_poi_fused_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_1(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = (xindex // 4)
tmp0 = tl.load(in_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp3 = triton_helpers.maximum(tmp1, tmp2)
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp7 = triton_helpers.maximum(tmp5, tmp6)
tmp8 = tmp0 - tmp7
tmp9 = tl_math.exp(tmp8)
tl.store(out_ptr0 + (x2), tmp9, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/kt/cktnex5febczl2ac6zugjmcksgsd5kjdufazv65vtepuwob3cb7a.py
# Topologically Sorted Source Nodes: [], Original ATen: []
# Source node to ATen node mapping:
# Graph fragment:
# %sum_dim_int_list : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%exp_default, [-1], True), kwargs = {})
# %div_tensor : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%exp_default, %sum_dim_int_list), kwargs = {})
# %eq_scalar : [num_users=1] = call_function[target=torch.ops.aten.eq.Scalar](args = (%view_default_2, -inf), kwargs = {})
# %logical_not_default : [num_users=1] = call_function[target=torch.ops.aten.logical_not.default](args = (%eq_scalar,), kwargs = {})
# %any_dim : [num_users=1] = call_function[target=torch.ops.aten.any.dim](args = (%logical_not_default, -1, True), kwargs = {})
# %logical_not_default_1 : [num_users=1] = call_function[target=torch.ops.aten.logical_not.default](args = (%any_dim,), kwargs = {})
# %full_default : [num_users=1] = call_function[target=torch.ops.aten.full.default](args = ([4, 4, 4, 4], 0), kwargs = {dtype: torch.float32, layout: torch.strided, device: cuda:0, pin_memory: False})
# %where_self : [num_users=2] = call_function[target=torch.ops.aten.where.self](args = (%logical_not_default_1, %full_default, %div_tensor), kwargs = {})
triton_poi_fused_2 = async_compile.triton('triton_poi_fused_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 9, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_2(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = (xindex // 4)
x2 = xindex
tmp0 = tl.load(in_ptr0 + (4*x1), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr0 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp18 = tl.load(in_ptr0 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp25 = tl.load(in_ptr1 + (x2), xmask)
tmp26 = tl.load(in_ptr1 + (4*x1), xmask, eviction_policy='evict_last')
tmp27 = tl.load(in_ptr1 + (1 + (4*x1)), xmask, eviction_policy='evict_last')
tmp29 = tl.load(in_ptr1 + (2 + (4*x1)), xmask, eviction_policy='evict_last')
tmp31 = tl.load(in_ptr1 + (3 + (4*x1)), xmask, eviction_policy='evict_last')
tmp1 = float("-inf")
tmp2 = tmp0 == tmp1
tmp3 = tmp2 == 0
tmp4 = tmp3.to(tl.int64)
tmp5 = (tmp4 != 0)
tmp7 = tmp6 == tmp1
tmp8 = tmp7 == 0
tmp9 = tmp8.to(tl.int64)
tmp10 = (tmp9 != 0)
tmp11 = tmp5 | tmp10
tmp13 = tmp12 == tmp1
tmp14 = tmp13 == 0
tmp15 = tmp14.to(tl.int64)
tmp16 = (tmp15 != 0)
tmp17 = tmp11 | tmp16
tmp19 = tmp18 == tmp1
tmp20 = tmp19 == 0
tmp21 = tmp20.to(tl.int64)
tmp22 = (tmp21 != 0)
tmp23 = tmp17 | tmp22
tmp24 = tmp23 == 0
tmp28 = tmp26 + tmp27
tmp30 = tmp28 + tmp29
tmp32 = tmp30 + tmp31
tmp33 = tmp25 / tmp32
tmp34 = 0.0
tmp35 = tl.where(tmp24, tmp34, tmp33)
tl.store(out_ptr0 + (x2), tmp35, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/vv/cvvnhithjvmvhfjufxwwzclfobkrgbyyteg66hp24r675f7elw4c.py
# Topologically Sorted Source Nodes: [], Original ATen: []
# Source node to ATen node mapping:
# Graph fragment:
# %clone_default_2 : [num_users=1] = call_function[target=torch.ops.aten.clone.default](args = (%expand_default_3,), kwargs = {memory_format: torch.contiguous_format})
triton_poi_fused_3 = async_compile.triton('triton_poi_fused_3', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16, 4], tile_hint=TileHint.DEFAULT,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_3', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_3(in_ptr0, in_ptr1, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = (yindex // 4)
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + (4*x2) + (16*y1)), xmask & ymask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + (y0), ymask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(out_ptr0 + (x2 + (4*y3)), tmp2, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/6t/c6t5a5ere3lqjiu7zh3uu4oxmpdoujdaqqmeunxqapgzo4m74uav.py
# Topologically Sorted Source Nodes: [context_layer_1], Original ATen: [aten.clone]
# Source node to ATen node mapping:
# context_layer_1 => clone_4
# Graph fragment:
# %clone_4 : [num_users=1] = call_function[target=torch.ops.aten.clone.default](args = (%permute_7,), kwargs = {memory_format: torch.contiguous_format})
triton_poi_fused_clone_4 = async_compile.triton('triton_poi_fused_clone_4', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16, 4], tile_hint=TileHint.SQUARE,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_clone_4', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_clone_4(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = (yindex // 4)
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + (4*x2) + (16*y1)), xmask & ymask, eviction_policy='evict_last')
tl.store(out_ptr0 + (x2 + (4*y3)), tmp0, xmask & ymask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8 = args
args.clear()
assert_size_stride(primals_1, (4, 4), (4, 1))
assert_size_stride(primals_2, (4, ), (1, ))
assert_size_stride(primals_3, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_4, (4, 4), (4, 1))
assert_size_stride(primals_5, (4, ), (1, ))
assert_size_stride(primals_6, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_7, (4, 4), (4, 1))
assert_size_stride(primals_8, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(primals_3, (16, 4), (4, 1), 0), reinterpret_tensor(primals_1, (4, 4), (1, 4), 0), out=buf0)
del primals_1
buf1 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(primals_6, (16, 4), (4, 1), 0), reinterpret_tensor(primals_4, (4, 4), (1, 4), 0), out=buf1)
del primals_4
buf2 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(primals_6, (16, 4), (4, 1), 0), reinterpret_tensor(primals_7, (4, 4), (1, 4), 0), out=buf2)
del primals_7
buf3 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
stream0 = get_raw_stream(0)
triton_poi_fused_0.run(buf0, primals_2, buf3, 16, 4, grid=grid(16, 4), stream=stream0)
del primals_2
buf4 = reinterpret_tensor(buf0, (4, 4, 1, 4), (16, 4, 4, 1), 0); del buf0 # reuse
# Topologically Sorted Source Nodes: [], Original ATen: []
triton_poi_fused_0.run(buf1, primals_5, buf4, 16, 4, grid=grid(16, 4), stream=stream0)
del primals_5
buf5 = empty_strided_cuda((16, 4, 4), (16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.bmm(reinterpret_tensor(buf3, (16, 4, 1), (4, 1, 0), 0), reinterpret_tensor(buf4, (16, 1, 4), (4, 0, 1), 0), out=buf5)
buf6 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
triton_poi_fused_1.run(buf5, buf6, 256, grid=grid(256), stream=stream0)
buf7 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
triton_poi_fused_2.run(buf5, buf6, buf7, 256, grid=grid(256), stream=stream0)
del buf5
del buf6
buf8 = reinterpret_tensor(buf1, (4, 4, 4, 1), (16, 4, 1, 1), 0); del buf1 # reuse
# Topologically Sorted Source Nodes: [], Original ATen: []
triton_poi_fused_3.run(buf2, primals_8, buf8, 16, 4, grid=grid(16, 4), stream=stream0)
del primals_8
buf9 = reinterpret_tensor(buf2, (16, 4, 1), (4, 1, 1), 0); del buf2 # reuse
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.bmm(reinterpret_tensor(buf7, (16, 4, 4), (16, 4, 1), 0), reinterpret_tensor(buf8, (16, 4, 1), (4, 1, 0), 0), out=buf9)
buf10 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 1), torch.float32)
# Topologically Sorted Source Nodes: [context_layer_1], Original ATen: [aten.clone]
triton_poi_fused_clone_4.run(buf9, buf10, 16, 4, grid=grid(16, 4), stream=stream0)
del buf9
return (reinterpret_tensor(buf10, (4, 4, 4), (16, 4, 1), 0), reinterpret_tensor(primals_3, (16, 4), (4, 1), 0), reinterpret_tensor(primals_6, (16, 4), (4, 1), 0), buf7, reinterpret_tensor(buf8, (16, 1, 4), (4, 1, 1), 0), reinterpret_tensor(buf3, (16, 1, 4), (4, 1, 1), 0), reinterpret_tensor(buf4, (16, 4, 1), (4, 1, 4), 0), )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((4, 4, 4), (16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((4, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_8 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| from _paritybench_helpers import _mock_config
import math
import torch
import torch.nn as nn
class BertAttention(nn.Module):
def __init__(self, config, ctx_dim=None):
super().__init__()
if config.hidden_size % config.num_attention_heads != 0:
raise ValueError(
'The hidden size (%d) is not a multiple of the number of attention heads (%d)'
% (config.hidden_size, config.num_attention_heads))
self.num_attention_heads = config.num_attention_heads
self.attention_head_size = int(config.hidden_size / config.
num_attention_heads)
self.all_head_size = (self.num_attention_heads * self.
attention_head_size)
if ctx_dim is None:
ctx_dim = config.hidden_size
self.query = nn.Linear(config.hidden_size, self.all_head_size)
self.key = nn.Linear(ctx_dim, self.all_head_size)
self.value = nn.Linear(ctx_dim, self.all_head_size)
self.dropout = nn.Dropout(config.attention_probs_dropout_prob)
def transpose_for_scores(self, x):
new_x_shape = x.size()[:-1] + (self.num_attention_heads, self.
attention_head_size)
x = x.view(*new_x_shape)
return x.permute(0, 2, 1, 3)
def forward(self, hidden_states, context, attention_mask=None):
mixed_query_layer = self.query(hidden_states)
mixed_key_layer = self.key(context)
mixed_value_layer = self.value(context)
query_layer = self.transpose_for_scores(mixed_query_layer)
key_layer = self.transpose_for_scores(mixed_key_layer)
value_layer = self.transpose_for_scores(mixed_value_layer)
attention_scores = torch.matmul(query_layer, key_layer.transpose(-1,
-2))
attention_scores = attention_scores / math.sqrt(self.
attention_head_size)
if attention_mask is not None:
attention_scores = attention_scores + attention_mask
attention_probs = nn.Softmax(dim=-1)(attention_scores)
attention_probs = self.dropout(attention_probs)
context_layer = torch.matmul(attention_probs, value_layer)
context_layer = context_layer.permute(0, 2, 1, 3).contiguous()
new_context_layer_shape = context_layer.size()[:-2] + (self.
all_head_size,)
context_layer = context_layer.view(*new_context_layer_shape)
return context_layer
def get_inputs():
return [torch.rand([4, 4, 4]), torch.rand([4, 4, 4])]
def get_init_inputs():
return [[], {'config': _mock_config(hidden_size=4, num_attention_heads=
4, attention_probs_dropout_prob=0.5)}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import math as tl_math
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_0(in_ptr0, in_ptr1, out_ptr0, ynumel, xnumel, YBLOCK:
tl.constexpr, XBLOCK: tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = yindex // 4
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + 4 * x2 + 16 * y1), xmask & ymask,
eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + y0, ymask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = 1.0
tmp4 = tmp2 * tmp3
tl.store(out_ptr0 + (x2 + 4 * y3), tmp4, xmask & ymask)
@triton.jit
def triton_poi_fused_1(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x1 = xindex // 4
tmp0 = tl.load(in_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last')
tmp3 = triton_helpers.maximum(tmp1, tmp2)
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp7 = triton_helpers.maximum(tmp5, tmp6)
tmp8 = tmp0 - tmp7
tmp9 = tl_math.exp(tmp8)
tl.store(out_ptr0 + x2, tmp9, xmask)
@triton.jit
def triton_poi_fused_2(in_ptr0, in_ptr1, out_ptr0, xnumel, XBLOCK: tl.constexpr
):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = xindex // 4
x2 = xindex
tmp0 = tl.load(in_ptr0 + 4 * x1, xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (1 + 4 * x1), xmask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr0 + (2 + 4 * x1), xmask, eviction_policy='evict_last'
)
tmp18 = tl.load(in_ptr0 + (3 + 4 * x1), xmask, eviction_policy='evict_last'
)
tmp25 = tl.load(in_ptr1 + x2, xmask)
tmp26 = tl.load(in_ptr1 + 4 * x1, xmask, eviction_policy='evict_last')
tmp27 = tl.load(in_ptr1 + (1 + 4 * x1), xmask, eviction_policy='evict_last'
)
tmp29 = tl.load(in_ptr1 + (2 + 4 * x1), xmask, eviction_policy='evict_last'
)
tmp31 = tl.load(in_ptr1 + (3 + 4 * x1), xmask, eviction_policy='evict_last'
)
tmp1 = float('-inf')
tmp2 = tmp0 == tmp1
tmp3 = tmp2 == 0
tmp4 = tmp3.to(tl.int64)
tmp5 = tmp4 != 0
tmp7 = tmp6 == tmp1
tmp8 = tmp7 == 0
tmp9 = tmp8.to(tl.int64)
tmp10 = tmp9 != 0
tmp11 = tmp5 | tmp10
tmp13 = tmp12 == tmp1
tmp14 = tmp13 == 0
tmp15 = tmp14.to(tl.int64)
tmp16 = tmp15 != 0
tmp17 = tmp11 | tmp16
tmp19 = tmp18 == tmp1
tmp20 = tmp19 == 0
tmp21 = tmp20.to(tl.int64)
tmp22 = tmp21 != 0
tmp23 = tmp17 | tmp22
tmp24 = tmp23 == 0
tmp28 = tmp26 + tmp27
tmp30 = tmp28 + tmp29
tmp32 = tmp30 + tmp31
tmp33 = tmp25 / tmp32
tmp34 = 0.0
tmp35 = tl.where(tmp24, tmp34, tmp33)
tl.store(out_ptr0 + x2, tmp35, xmask)
@triton.jit
def triton_poi_fused_3(in_ptr0, in_ptr1, out_ptr0, ynumel, xnumel, YBLOCK:
tl.constexpr, XBLOCK: tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = yindex // 4
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + 4 * x2 + 16 * y1), xmask & ymask,
eviction_policy='evict_last')
tmp1 = tl.load(in_ptr1 + y0, ymask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(out_ptr0 + (x2 + 4 * y3), tmp2, xmask & ymask)
@triton.jit
def triton_poi_fused_clone_4(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK: tl.
constexpr, XBLOCK: tl.constexpr):
ynumel = 16
xnumel = 4
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x2 = xindex
y0 = yindex % 4
y1 = yindex // 4
y3 = yindex
tmp0 = tl.load(in_ptr0 + (y0 + 4 * x2 + 16 * y1), xmask & ymask,
eviction_policy='evict_last')
tl.store(out_ptr0 + (x2 + 4 * y3), tmp0, xmask & ymask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7, primals_8) = args
args.clear()
assert_size_stride(primals_1, (4, 4), (4, 1))
assert_size_stride(primals_2, (4,), (1,))
assert_size_stride(primals_3, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_4, (4, 4), (4, 1))
assert_size_stride(primals_5, (4,), (1,))
assert_size_stride(primals_6, (4, 4, 4), (16, 4, 1))
assert_size_stride(primals_7, (4, 4), (4, 1))
assert_size_stride(primals_8, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_3, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_1, (4, 4), (1, 4), 0), out=buf0)
del primals_1
buf1 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_6, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_4, (4, 4), (1, 4), 0), out=buf1)
del primals_4
buf2 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_6, (16, 4), (4, 1), 0),
reinterpret_tensor(primals_7, (4, 4), (1, 4), 0), out=buf2)
del primals_7
buf3 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_0[grid(16, 4)](buf0, primals_2, buf3, 16, 4,
XBLOCK=4, YBLOCK=8, num_warps=1, num_stages=1)
del primals_2
buf4 = reinterpret_tensor(buf0, (4, 4, 1, 4), (16, 4, 4, 1), 0)
del buf0
triton_poi_fused_0[grid(16, 4)](buf1, primals_5, buf4, 16, 4,
XBLOCK=4, YBLOCK=8, num_warps=1, num_stages=1)
del primals_5
buf5 = empty_strided_cuda((16, 4, 4), (16, 4, 1), torch.float32)
extern_kernels.bmm(reinterpret_tensor(buf3, (16, 4, 1), (4, 1, 0),
0), reinterpret_tensor(buf4, (16, 1, 4), (4, 0, 1), 0), out=buf5)
buf6 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
triton_poi_fused_1[grid(256)](buf5, buf6, 256, XBLOCK=128,
num_warps=4, num_stages=1)
buf7 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
triton_poi_fused_2[grid(256)](buf5, buf6, buf7, 256, XBLOCK=128,
num_warps=4, num_stages=1)
del buf5
del buf6
buf8 = reinterpret_tensor(buf1, (4, 4, 4, 1), (16, 4, 1, 1), 0)
del buf1
triton_poi_fused_3[grid(16, 4)](buf2, primals_8, buf8, 16, 4,
XBLOCK=4, YBLOCK=16, num_warps=1, num_stages=1)
del primals_8
buf9 = reinterpret_tensor(buf2, (16, 4, 1), (4, 1, 1), 0)
del buf2
extern_kernels.bmm(reinterpret_tensor(buf7, (16, 4, 4), (16, 4, 1),
0), reinterpret_tensor(buf8, (16, 4, 1), (4, 1, 0), 0), out=buf9)
buf10 = empty_strided_cuda((4, 4, 4, 1), (16, 4, 1, 1), torch.float32)
triton_poi_fused_clone_4[grid(16, 4)](buf9, buf10, 16, 4, XBLOCK=4,
YBLOCK=16, num_warps=1, num_stages=1)
del buf9
return reinterpret_tensor(buf10, (4, 4, 4), (16, 4, 1), 0
), reinterpret_tensor(primals_3, (16, 4), (4, 1), 0
), reinterpret_tensor(primals_6, (16, 4), (4, 1), 0
), buf7, reinterpret_tensor(buf8, (16, 1, 4), (4, 1, 1), 0
), reinterpret_tensor(buf3, (16, 1, 4), (4, 1, 1), 0
), reinterpret_tensor(buf4, (16, 4, 1), (4, 1, 4), 0)
class BertAttentionNew(nn.Module):
def __init__(self, config, ctx_dim=None):
super().__init__()
if config.hidden_size % config.num_attention_heads != 0:
raise ValueError(
'The hidden size (%d) is not a multiple of the number of attention heads (%d)'
% (config.hidden_size, config.num_attention_heads))
self.num_attention_heads = config.num_attention_heads
self.attention_head_size = int(config.hidden_size / config.
num_attention_heads)
self.all_head_size = (self.num_attention_heads * self.
attention_head_size)
if ctx_dim is None:
ctx_dim = config.hidden_size
self.query = nn.Linear(config.hidden_size, self.all_head_size)
self.key = nn.Linear(ctx_dim, self.all_head_size)
self.value = nn.Linear(ctx_dim, self.all_head_size)
self.dropout = nn.Dropout(config.attention_probs_dropout_prob)
def transpose_for_scores(self, x):
new_x_shape = x.size()[:-1] + (self.num_attention_heads, self.
attention_head_size)
x = x.view(*new_x_shape)
return x.permute(0, 2, 1, 3)
def forward(self, input_0, input_1):
primals_1 = self.query.weight
primals_2 = self.query.bias
primals_4 = self.key.weight
primals_5 = self.key.bias
primals_7 = self.value.weight
primals_8 = self.value.bias
primals_3 = input_0
primals_6 = input_1
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7, primals_8])
return output[0]
| AsmitaBhat30/lxmert | BertAttention | false | 6,921 | [
"MIT"
] | 1 | 90292dc36a25c04c4f76fe9119e3141d5dc05874 | https://github.com/AsmitaBhat30/lxmert/tree/90292dc36a25c04c4f76fe9119e3141d5dc05874 | from _paritybench_helpers import _mock_config
import math
import torch
import torch.nn as nn
class Model(nn.Module):
def __init__(self, config, ctx_dim=None):
super().__init__()
if config.hidden_size % config.num_attention_heads != 0:
raise ValueError(
'The hidden size (%d) is not a multiple of the number of attention heads (%d)'
% (config.hidden_size, config.num_attention_heads))
self.num_attention_heads = config.num_attention_heads
self.attention_head_size = int(config.hidden_size / config.
num_attention_heads)
self.all_head_size = (self.num_attention_heads * self.
attention_head_size)
if ctx_dim is None:
ctx_dim = config.hidden_size
self.query = nn.Linear(config.hidden_size, self.all_head_size)
self.key = nn.Linear(ctx_dim, self.all_head_size)
self.value = nn.Linear(ctx_dim, self.all_head_size)
self.dropout = nn.Dropout(config.attention_probs_dropout_prob)
def transpose_for_scores(self, x):
new_x_shape = x.size()[:-1] + (self.num_attention_heads, self.
attention_head_size)
x = x.view(*new_x_shape)
return x.permute(0, 2, 1, 3)
def forward(self, hidden_states, context, attention_mask=None):
mixed_query_layer = self.query(hidden_states)
mixed_key_layer = self.key(context)
mixed_value_layer = self.value(context)
query_layer = self.transpose_for_scores(mixed_query_layer)
key_layer = self.transpose_for_scores(mixed_key_layer)
value_layer = self.transpose_for_scores(mixed_value_layer)
attention_scores = torch.matmul(query_layer, key_layer.transpose(-1,
-2))
attention_scores = attention_scores / math.sqrt(self.
attention_head_size)
if attention_mask is not None:
attention_scores = attention_scores + attention_mask
attention_probs = nn.Softmax(dim=-1)(attention_scores)
attention_probs = self.dropout(attention_probs)
context_layer = torch.matmul(attention_probs, value_layer)
context_layer = context_layer.permute(0, 2, 1, 3).contiguous()
new_context_layer_shape = context_layer.size()[:-2] + (self.
all_head_size,)
context_layer = context_layer.view(*new_context_layer_shape)
return context_layer
def get_inputs():
return [torch.rand([4, 4, 4]), torch.rand([4, 4, 4])]
def get_init_inputs():
return [[], {'config': _mock_config(hidden_size=4, num_attention_heads=
4, attention_probs_dropout_prob=0.5)}]
|
TVLoss | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/ny/cnyr7377i6636kt4c5ujuskifpyrhu7b5eo3e24h2zgxlygwk7wp.py
# Topologically Sorted Source Nodes: [x_diff, pow_1, y_diff, pow_2, add, mean], Original ATen: [aten.sub, aten.pow, aten.add, aten.mean]
# Source node to ATen node mapping:
# add => add
# mean => mean
# pow_1 => pow_1
# pow_2 => pow_2
# x_diff => sub
# y_diff => sub_1
# Graph fragment:
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%slice_2, %slice_4), kwargs = {})
# %pow_1 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%sub, 2), kwargs = {})
# %sub_1 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%slice_6, %slice_8), kwargs = {})
# %pow_2 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%sub_1, 2), kwargs = {})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%pow_1, %pow_2), kwargs = {})
# %mean : [num_users=1] = call_function[target=torch.ops.aten.mean.default](args = (%add,), kwargs = {})
triton_per_fused_add_mean_pow_sub_0 = async_compile.triton('triton_per_fused_add_mean_pow_sub_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[1, 256],
reduction_hint=ReductionHint.INNER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {2: 1}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 3), equal_to_1=(2,))]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused_add_mean_pow_sub_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': True, 'num_load': 3, 'num_reduction': 1, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused_add_mean_pow_sub_0(in_out_ptr0, in_ptr0, xnumel, rnumel):
xnumel = 1
XBLOCK: tl.constexpr = 1
rnumel = 256
RBLOCK: tl.constexpr = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = tl.full([1], xoffset, tl.int32)
xmask = tl.full([RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[:]
roffset = 0
rmask = tl.full([RBLOCK], True, tl.int1)
r0 = rindex % 4
r1 = (rindex // 4) % 4
r2 = (rindex // 16)
tmp0 = tl.load(in_ptr0 + ((4*((3) * ((3) <= (r1)) + (r1) * ((r1) < (3)))) + (16*r2) + ((3) * ((3) <= (1 + r0)) + (1 + r0) * ((1 + r0) < (3)))), None)
tmp1 = tl.load(in_ptr0 + ((4*((3) * ((3) <= (r1)) + (r1) * ((r1) < (3)))) + (16*r2) + ((3) * ((3) <= (r0)) + (r0) * ((r0) < (3)))), None)
tmp4 = tl.load(in_ptr0 + ((4*((3) * ((3) <= (1 + r1)) + (1 + r1) * ((1 + r1) < (3)))) + (16*r2) + ((3) * ((3) <= (r0)) + (r0) * ((r0) < (3)))), None)
tmp2 = tmp0 - tmp1
tmp3 = tmp2 * tmp2
tmp5 = tmp4 - tmp1
tmp6 = tmp5 * tmp5
tmp7 = tmp3 + tmp6
tmp8 = tl.broadcast_to(tmp7, [RBLOCK])
tmp10 = triton_helpers.promote_to_tensor(tl.sum(tmp8, 0))
tmp11 = 256.0
tmp12 = tmp10 / tmp11
tl.debug_barrier()
tl.store(in_out_ptr0 + (tl.full([1], 0, tl.int32)), tmp12, None)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((), (), torch.float32)
buf1 = buf0; del buf0 # reuse
# Topologically Sorted Source Nodes: [x_diff, pow_1, y_diff, pow_2, add, mean], Original ATen: [aten.sub, aten.pow, aten.add, aten.mean]
stream0 = get_raw_stream(0)
triton_per_fused_add_mean_pow_sub_0.run(buf1, arg0_1, 1, 256, grid=grid(1), stream=stream0)
del arg0_1
return (buf1, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([arg0_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.nn.functional as F
class TVLoss(nn.Module):
"""L2 total variation loss, as in Mahendran et al."""
def forward(self, input):
input = F.pad(input, (0, 1, 0, 1), 'replicate')
x_diff = input[..., :-1, 1:] - input[..., :-1, :-1]
y_diff = input[..., 1:, :-1] - input[..., :-1, :-1]
return (x_diff ** 2 + y_diff ** 2).mean()
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {}]
| import torch
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_per_fused_add_mean_pow_sub_0(in_out_ptr0, in_ptr0, xnumel, rnumel):
XBLOCK: tl.constexpr = 1
RBLOCK: tl.constexpr = 256
xoffset = tl.program_id(0) * XBLOCK
tl.full([1], xoffset, tl.int32)
tl.full([RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[:]
tl.full([RBLOCK], True, tl.int1)
r0 = rindex % 4
r1 = rindex // 4 % 4
r2 = rindex // 16
tmp0 = tl.load(in_ptr0 + (4 * (3 * (3 <= r1) + r1 * (r1 < 3)) + 16 * r2 +
(3 * (3 <= 1 + r0) + (1 + r0) * (1 + r0 < 3))), None)
tmp1 = tl.load(in_ptr0 + (4 * (3 * (3 <= r1) + r1 * (r1 < 3)) + 16 * r2 +
(3 * (3 <= r0) + r0 * (r0 < 3))), None)
tmp4 = tl.load(in_ptr0 + (4 * (3 * (3 <= 1 + r1) + (1 + r1) * (1 + r1 <
3)) + 16 * r2 + (3 * (3 <= r0) + r0 * (r0 < 3))), None)
tmp2 = tmp0 - tmp1
tmp3 = tmp2 * tmp2
tmp5 = tmp4 - tmp1
tmp6 = tmp5 * tmp5
tmp7 = tmp3 + tmp6
tmp8 = tl.broadcast_to(tmp7, [RBLOCK])
tmp10 = triton_helpers.promote_to_tensor(tl.sum(tmp8, 0))
tmp11 = 256.0
tmp12 = tmp10 / tmp11
tl.debug_barrier()
tl.store(in_out_ptr0 + tl.full([1], 0, tl.int32), tmp12, None)
def call(args):
arg0_1, = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((), (), torch.float32)
buf1 = buf0
del buf0
get_raw_stream(0)
triton_per_fused_add_mean_pow_sub_0[grid(1)](buf1, arg0_1, 1, 256,
num_warps=2, num_stages=1)
del arg0_1
return buf1,
class TVLossNew(nn.Module):
"""L2 total variation loss, as in Mahendran et al."""
def forward(self, input_0):
arg0_1 = input_0
output = call([arg0_1])
return output[0]
| jayChung0302/SideProject-sudalchongbo | TVLoss | false | 6,922 | [
"MIT"
] | 1 | fb0a3d0aee53ba24d3b8ec2dd8c52d0e8f6c33d7 | https://github.com/jayChung0302/SideProject-sudalchongbo/tree/fb0a3d0aee53ba24d3b8ec2dd8c52d0e8f6c33d7 | import torch
import torch.nn as nn
import torch.nn.functional as F
class Model(nn.Module):
"""L2 total variation loss, as in Mahendran et al."""
def forward(self, input):
input = F.pad(input, (0, 1, 0, 1), 'replicate')
x_diff = input[..., :-1, 1:] - input[..., :-1, :-1]
y_diff = input[..., 1:, :-1] - input[..., :-1, :-1]
return (x_diff ** 2 + y_diff ** 2).mean()
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return []
|
FocalLoss | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/hh/chhirykkudhsj4gqr5bunmner6jehbnsqvm6c5yuiwe7zimguzmq.py
# Topologically Sorted Source Nodes: [eq_1, probs, sub_1, probs_1_gamma, ge, softplus, softplus_1, sub, log_probs, term1, probs_gamma, ge_1, neg, softplus_2, add, softplus_3, neg_1, log_1_probs, term2, where_2, coeff, setitem, mul_, loss, loss_1], Original ATen: [aten.eq, aten.sigmoid, aten.rsub, aten.pow, aten.ge, aten.softplus, aten.sub, aten.where, aten.mul, aten.neg, aten.add, aten.fill, aten.lift_fresh, aten.index_put, aten.mean]
# Source node to ATen node mapping:
# add => add
# coeff => full_default
# eq_1 => eq_1
# ge => ge
# ge_1 => ge_1
# log_1_probs => where_5
# log_probs => where_2
# loss => neg_2
# loss_1 => mean
# mul_ => mul_6
# neg => neg
# neg_1 => neg_1
# probs => sigmoid
# probs_1_gamma => pow_2
# probs_gamma => pow_1
# setitem => full_default_1, index_put
# softplus => div, exp, gt, log1p, mul, where
# softplus_1 => div_1, exp_1, gt_1, log1p_1, mul_1, where_1
# softplus_2 => div_2, exp_2, gt_2, log1p_2, mul_2, where_3
# softplus_3 => div_3, exp_3, gt_3, log1p_3, mul_3, where_4
# sub => sub
# sub_1 => sub_1
# term1 => mul_4
# term2 => mul_5
# where_2 => where_6
# Graph fragment:
# %eq_1 : [num_users=1] = call_function[target=torch.ops.aten.eq.Scalar](args = (%arg1_1, 1), kwargs = {})
# %sigmoid : [num_users=2] = call_function[target=torch.ops.aten.sigmoid.default](args = (%arg0_1,), kwargs = {})
# %sub_1 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (1.0, %sigmoid), kwargs = {})
# %pow_2 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%sub_1, 2), kwargs = {})
# %ge : [num_users=1] = call_function[target=torch.ops.aten.ge.Scalar](args = (%arg0_1, 0), kwargs = {})
# %mul : [num_users=2] = call_function[target=torch.ops.aten.mul.Tensor](args = (%arg0_1, -1), kwargs = {})
# %gt : [num_users=1] = call_function[target=torch.ops.aten.gt.Scalar](args = (%mul, 50), kwargs = {})
# %exp : [num_users=1] = call_function[target=torch.ops.aten.exp.default](args = (%mul,), kwargs = {})
# %log1p : [num_users=1] = call_function[target=torch.ops.aten.log1p.default](args = (%exp,), kwargs = {})
# %div : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%log1p, -1), kwargs = {})
# %where : [num_users=1] = call_function[target=torch.ops.aten.where.self](args = (%gt, %arg0_1, %div), kwargs = {})
# %mul_1 : [num_users=2] = call_function[target=torch.ops.aten.mul.Tensor](args = (%arg0_1, 1), kwargs = {})
# %gt_1 : [num_users=1] = call_function[target=torch.ops.aten.gt.Scalar](args = (%mul_1, 50), kwargs = {})
# %exp_1 : [num_users=1] = call_function[target=torch.ops.aten.exp.default](args = (%mul_1,), kwargs = {})
# %log1p_1 : [num_users=1] = call_function[target=torch.ops.aten.log1p.default](args = (%exp_1,), kwargs = {})
# %div_1 : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%log1p_1, 1), kwargs = {})
# %where_1 : [num_users=1] = call_function[target=torch.ops.aten.where.self](args = (%gt_1, %arg0_1, %div_1), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%arg0_1, %where_1), kwargs = {})
# %where_2 : [num_users=1] = call_function[target=torch.ops.aten.where.self](args = (%ge, %where, %sub), kwargs = {})
# %mul_4 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%pow_2, %where_2), kwargs = {})
# %pow_1 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%sigmoid, 2), kwargs = {})
# %ge_1 : [num_users=1] = call_function[target=torch.ops.aten.ge.Scalar](args = (%arg0_1, 0), kwargs = {})
# %neg : [num_users=1] = call_function[target=torch.ops.aten.neg.default](args = (%arg0_1,), kwargs = {})
# %mul_2 : [num_users=2] = call_function[target=torch.ops.aten.mul.Tensor](args = (%arg0_1, -1), kwargs = {})
# %gt_2 : [num_users=1] = call_function[target=torch.ops.aten.gt.Scalar](args = (%mul_2, 50), kwargs = {})
# %exp_2 : [num_users=1] = call_function[target=torch.ops.aten.exp.default](args = (%mul_2,), kwargs = {})
# %log1p_2 : [num_users=1] = call_function[target=torch.ops.aten.log1p.default](args = (%exp_2,), kwargs = {})
# %div_2 : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%log1p_2, -1), kwargs = {})
# %where_3 : [num_users=1] = call_function[target=torch.ops.aten.where.self](args = (%gt_2, %arg0_1, %div_2), kwargs = {})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%neg, %where_3), kwargs = {})
# %mul_3 : [num_users=2] = call_function[target=torch.ops.aten.mul.Tensor](args = (%arg0_1, 1), kwargs = {})
# %gt_3 : [num_users=1] = call_function[target=torch.ops.aten.gt.Scalar](args = (%mul_3, 50), kwargs = {})
# %exp_3 : [num_users=1] = call_function[target=torch.ops.aten.exp.default](args = (%mul_3,), kwargs = {})
# %log1p_3 : [num_users=1] = call_function[target=torch.ops.aten.log1p.default](args = (%exp_3,), kwargs = {})
# %div_3 : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%log1p_3, 1), kwargs = {})
# %where_4 : [num_users=1] = call_function[target=torch.ops.aten.where.self](args = (%gt_3, %arg0_1, %div_3), kwargs = {})
# %neg_1 : [num_users=1] = call_function[target=torch.ops.aten.neg.default](args = (%where_4,), kwargs = {})
# %where_5 : [num_users=1] = call_function[target=torch.ops.aten.where.self](args = (%ge_1, %add, %neg_1), kwargs = {})
# %mul_5 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%pow_1, %where_5), kwargs = {})
# %where_6 : [num_users=1] = call_function[target=torch.ops.aten.where.self](args = (%eq_1, %mul_4, %mul_5), kwargs = {})
# %full_default : [num_users=1] = call_function[target=torch.ops.aten.full.default](args = ([4, 4, 4, 4], 0), kwargs = {dtype: torch.float32, layout: torch.strided, device: cuda:0, pin_memory: False})
# %full_default_1 : [num_users=1] = call_function[target=torch.ops.aten.full.default](args = ([], 1.0), kwargs = {dtype: torch.float32, layout: torch.strided, device: cpu, pin_memory: False})
# %index_put : [num_users=1] = call_function[target=torch.ops.aten.index_put_.default](args = (%full_default, [%eq], %full_default_1), kwargs = {})
# %mul_6 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%where_6, %index_put), kwargs = {})
# %neg_2 : [num_users=1] = call_function[target=torch.ops.aten.neg.default](args = (%mul_6,), kwargs = {})
# %mean : [num_users=1] = call_function[target=torch.ops.aten.mean.default](args = (%neg_2,), kwargs = {})
triton_per_fused_add_eq_fill_ge_index_put_lift_fresh_mean_mul_neg_pow_rsub_sigmoid_softplus_sub_where_0 = async_compile.triton('triton_per_fused_add_eq_fill_ge_index_put_lift_fresh_mean_mul_neg_pow_rsub_sigmoid_softplus_sub_where_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[1, 256],
reduction_hint=ReductionHint.INNER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: 'i32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {3: 1}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 4), equal_to_1=(3,))]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused_add_eq_fill_ge_index_put_lift_fresh_mean_mul_neg_pow_rsub_sigmoid_softplus_sub_where_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': True, 'num_load': 2, 'num_reduction': 1, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused_add_eq_fill_ge_index_put_lift_fresh_mean_mul_neg_pow_rsub_sigmoid_softplus_sub_where_0(in_out_ptr0, in_ptr0, in_ptr1, xnumel, rnumel):
xnumel = 1
XBLOCK: tl.constexpr = 1
rnumel = 256
RBLOCK: tl.constexpr = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = tl.full([1], xoffset, tl.int32)
xmask = tl.full([RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[:]
roffset = 0
rmask = tl.full([RBLOCK], True, tl.int1)
r0 = rindex
tmp0 = tl.load(in_ptr0 + (r0), None)
tmp3 = tl.load(in_ptr1 + (r0), None)
tmp1 = 1.0
tmp2 = tmp0 == tmp1
tmp4 = tl.sigmoid(tmp3)
tmp5 = tmp1 - tmp4
tmp6 = tmp5 * tmp5
tmp7 = 0.0
tmp8 = tmp3 >= tmp7
tmp9 = -1.0
tmp10 = tmp3 * tmp9
tmp11 = 50.0
tmp12 = tmp10 > tmp11
tmp13 = tl_math.exp(tmp10)
tmp14 = libdevice.log1p(tmp13)
tmp15 = tmp14 * tmp9
tmp16 = tl.where(tmp12, tmp3, tmp15)
tmp17 = tmp3 * tmp1
tmp18 = tmp17 > tmp11
tmp19 = tl_math.exp(tmp17)
tmp20 = libdevice.log1p(tmp19)
tmp21 = tmp20 * tmp1
tmp22 = tl.where(tmp18, tmp3, tmp21)
tmp23 = tmp3 - tmp22
tmp24 = tl.where(tmp8, tmp16, tmp23)
tmp25 = tmp6 * tmp24
tmp26 = tmp4 * tmp4
tmp27 = -tmp3
tmp28 = tmp27 + tmp16
tmp29 = -tmp22
tmp30 = tl.where(tmp8, tmp28, tmp29)
tmp31 = tmp26 * tmp30
tmp32 = tl.where(tmp2, tmp25, tmp31)
tmp33 = tl.where(tmp2, tmp1, tmp7)
tmp34 = tmp32 * tmp33
tmp35 = -tmp34
tmp36 = tl.broadcast_to(tmp35, [RBLOCK])
tmp38 = triton_helpers.promote_to_tensor(tl.sum(tmp36, 0))
tmp39 = 256.0
tmp40 = tmp38 / tmp39
tl.debug_barrier()
tl.store(in_out_ptr0 + (tl.full([1], 0, tl.int32)), tmp40, None)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, arg1_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg1_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf2 = empty_strided_cuda((), (), torch.float32)
buf3 = buf2; del buf2 # reuse
# Topologically Sorted Source Nodes: [eq_1, probs, sub_1, probs_1_gamma, ge, softplus, softplus_1, sub, log_probs, term1, probs_gamma, ge_1, neg, softplus_2, add, softplus_3, neg_1, log_1_probs, term2, where_2, coeff, setitem, mul_, loss, loss_1], Original ATen: [aten.eq, aten.sigmoid, aten.rsub, aten.pow, aten.ge, aten.softplus, aten.sub, aten.where, aten.mul, aten.neg, aten.add, aten.fill, aten.lift_fresh, aten.index_put, aten.mean]
stream0 = get_raw_stream(0)
triton_per_fused_add_eq_fill_ge_index_put_lift_fresh_mean_mul_neg_pow_rsub_sigmoid_softplus_sub_where_0.run(buf3, arg1_1, arg0_1, 1, 256, grid=grid(1), stream=stream0)
del arg0_1
del arg1_1
return (buf3, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
arg1_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([arg0_1, arg1_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.nn.functional as F
class FocalSigmoidLossFunc(torch.autograd.Function):
"""
compute backward directly for better numeric stability
"""
@staticmethod
def forward(ctx, logits, label, alpha, gamma):
logits = logits.float()
coeff = torch.empty_like(logits).fill_(1 - alpha)
coeff[label == 1] = alpha
probs = torch.sigmoid(logits)
log_probs = torch.where(logits >= 0, F.softplus(logits, -1, 50),
logits - F.softplus(logits, 1, 50))
log_1_probs = torch.where(logits >= 0, -logits + F.softplus(logits,
-1, 50), -F.softplus(logits, 1, 50))
probs_gamma = probs ** gamma
probs_1_gamma = (1.0 - probs) ** gamma
ctx.coeff = coeff
ctx.probs = probs
ctx.log_probs = log_probs
ctx.log_1_probs = log_1_probs
ctx.probs_gamma = probs_gamma
ctx.probs_1_gamma = probs_1_gamma
ctx.label = label
ctx.gamma = gamma
term1 = probs_1_gamma * log_probs
term2 = probs_gamma * log_1_probs
loss = torch.where(label == 1, term1, term2).mul_(coeff).neg_()
return loss
@staticmethod
def backward(ctx, grad_output):
"""
compute gradient of focal loss
"""
coeff = ctx.coeff
probs = ctx.probs
log_probs = ctx.log_probs
log_1_probs = ctx.log_1_probs
probs_gamma = ctx.probs_gamma
probs_1_gamma = ctx.probs_1_gamma
label = ctx.label
gamma = ctx.gamma
term1 = (1.0 - probs - gamma * probs * log_probs).mul_(probs_1_gamma
).neg_()
term2 = (probs - gamma * (1.0 - probs) * log_1_probs).mul_(probs_gamma)
grads = torch.where(label == 1, term1, term2).mul_(coeff).mul_(
grad_output)
return grads, None, None, None
class FocalLoss(nn.Module):
"""
This use better formula to compute the gradient, which has better numeric stability
"""
def __init__(self, alpha=1, gamma=2, reduction='mean'):
super().__init__()
self.alpha = alpha
self.gamma = gamma
self.reduction = reduction
def forward(self, logits, label):
loss = FocalSigmoidLossFunc.apply(logits, label, self.alpha, self.gamma
)
if self.reduction == 'mean':
loss = loss.mean()
if self.reduction == 'sum':
loss = loss.sum()
return loss
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {}]
| import torch
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
import torch.nn as nn
import torch.nn.functional as F
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_per_fused_add_eq_fill_ge_index_put_lift_fresh_mean_mul_neg_pow_rsub_sigmoid_softplus_sub_where_0(
in_out_ptr0, in_ptr0, in_ptr1, xnumel, rnumel):
XBLOCK: tl.constexpr = 1
RBLOCK: tl.constexpr = 256
xoffset = tl.program_id(0) * XBLOCK
tl.full([1], xoffset, tl.int32)
tl.full([RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[:]
tl.full([RBLOCK], True, tl.int1)
r0 = rindex
tmp0 = tl.load(in_ptr0 + r0, None)
tmp3 = tl.load(in_ptr1 + r0, None)
tmp1 = 1.0
tmp2 = tmp0 == tmp1
tmp4 = tl.sigmoid(tmp3)
tmp5 = tmp1 - tmp4
tmp6 = tmp5 * tmp5
tmp7 = 0.0
tmp8 = tmp3 >= tmp7
tmp9 = -1.0
tmp10 = tmp3 * tmp9
tmp11 = 50.0
tmp12 = tmp10 > tmp11
tmp13 = tl_math.exp(tmp10)
tmp14 = libdevice.log1p(tmp13)
tmp15 = tmp14 * tmp9
tmp16 = tl.where(tmp12, tmp3, tmp15)
tmp17 = tmp3 * tmp1
tmp18 = tmp17 > tmp11
tmp19 = tl_math.exp(tmp17)
tmp20 = libdevice.log1p(tmp19)
tmp21 = tmp20 * tmp1
tmp22 = tl.where(tmp18, tmp3, tmp21)
tmp23 = tmp3 - tmp22
tmp24 = tl.where(tmp8, tmp16, tmp23)
tmp25 = tmp6 * tmp24
tmp26 = tmp4 * tmp4
tmp27 = -tmp3
tmp28 = tmp27 + tmp16
tmp29 = -tmp22
tmp30 = tl.where(tmp8, tmp28, tmp29)
tmp31 = tmp26 * tmp30
tmp32 = tl.where(tmp2, tmp25, tmp31)
tmp33 = tl.where(tmp2, tmp1, tmp7)
tmp34 = tmp32 * tmp33
tmp35 = -tmp34
tmp36 = tl.broadcast_to(tmp35, [RBLOCK])
tmp38 = triton_helpers.promote_to_tensor(tl.sum(tmp36, 0))
tmp39 = 256.0
tmp40 = tmp38 / tmp39
tl.debug_barrier()
tl.store(in_out_ptr0 + tl.full([1], 0, tl.int32), tmp40, None)
def call(args):
arg0_1, arg1_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg1_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf2 = empty_strided_cuda((), (), torch.float32)
buf3 = buf2
del buf2
get_raw_stream(0)
triton_per_fused_add_eq_fill_ge_index_put_lift_fresh_mean_mul_neg_pow_rsub_sigmoid_softplus_sub_where_0[
grid(1)](buf3, arg1_1, arg0_1, 1, 256, num_warps=2, num_stages=1)
del arg0_1
del arg1_1
return buf3,
class FocalSigmoidLossFunc(torch.autograd.Function):
"""
compute backward directly for better numeric stability
"""
@staticmethod
def forward(ctx, logits, label, alpha, gamma):
logits = logits.float()
coeff = torch.empty_like(logits).fill_(1 - alpha)
coeff[label == 1] = alpha
probs = torch.sigmoid(logits)
log_probs = torch.where(logits >= 0, F.softplus(logits, -1, 50),
logits - F.softplus(logits, 1, 50))
log_1_probs = torch.where(logits >= 0, -logits + F.softplus(logits,
-1, 50), -F.softplus(logits, 1, 50))
probs_gamma = probs ** gamma
probs_1_gamma = (1.0 - probs) ** gamma
ctx.coeff = coeff
ctx.probs = probs
ctx.log_probs = log_probs
ctx.log_1_probs = log_1_probs
ctx.probs_gamma = probs_gamma
ctx.probs_1_gamma = probs_1_gamma
ctx.label = label
ctx.gamma = gamma
term1 = probs_1_gamma * log_probs
term2 = probs_gamma * log_1_probs
loss = torch.where(label == 1, term1, term2).mul_(coeff).neg_()
return loss
@staticmethod
def backward(ctx, grad_output):
"""
compute gradient of focal loss
"""
coeff = ctx.coeff
probs = ctx.probs
log_probs = ctx.log_probs
log_1_probs = ctx.log_1_probs
probs_gamma = ctx.probs_gamma
probs_1_gamma = ctx.probs_1_gamma
label = ctx.label
gamma = ctx.gamma
term1 = (1.0 - probs - gamma * probs * log_probs).mul_(probs_1_gamma
).neg_()
term2 = (probs - gamma * (1.0 - probs) * log_1_probs).mul_(probs_gamma)
grads = torch.where(label == 1, term1, term2).mul_(coeff).mul_(
grad_output)
return grads, None, None, None
class FocalLossNew(nn.Module):
"""
This use better formula to compute the gradient, which has better numeric stability
"""
def __init__(self, alpha=1, gamma=2, reduction='mean'):
super().__init__()
self.alpha = alpha
self.gamma = gamma
self.reduction = reduction
def forward(self, input_0, input_1):
arg0_1 = input_0
arg1_1 = input_1
output = call([arg0_1, arg1_1])
return output[0]
| jaredaevans/UltrafastNST | FocalLoss | false | 6,923 | [
"MIT"
] | 1 | 6671c6b618ce6bb4920b15f782be962e484a5423 | https://github.com/jaredaevans/UltrafastNST/tree/6671c6b618ce6bb4920b15f782be962e484a5423 | import torch
import torch.nn as nn
import torch.nn.functional as F
class FocalSigmoidLossFunc(torch.autograd.Function):
"""
compute backward directly for better numeric stability
"""
@staticmethod
def forward(ctx, logits, label, alpha, gamma):
logits = logits.float()
coeff = torch.empty_like(logits).fill_(1 - alpha)
coeff[label == 1] = alpha
probs = torch.sigmoid(logits)
log_probs = torch.where(logits >= 0, F.softplus(logits, -1, 50),
logits - F.softplus(logits, 1, 50))
log_1_probs = torch.where(logits >= 0, -logits + F.softplus(logits,
-1, 50), -F.softplus(logits, 1, 50))
probs_gamma = probs ** gamma
probs_1_gamma = (1.0 - probs) ** gamma
ctx.coeff = coeff
ctx.probs = probs
ctx.log_probs = log_probs
ctx.log_1_probs = log_1_probs
ctx.probs_gamma = probs_gamma
ctx.probs_1_gamma = probs_1_gamma
ctx.label = label
ctx.gamma = gamma
term1 = probs_1_gamma * log_probs
term2 = probs_gamma * log_1_probs
loss = torch.where(label == 1, term1, term2).mul_(coeff).neg_()
return loss
@staticmethod
def backward(ctx, grad_output):
"""
compute gradient of focal loss
"""
coeff = ctx.coeff
probs = ctx.probs
log_probs = ctx.log_probs
log_1_probs = ctx.log_1_probs
probs_gamma = ctx.probs_gamma
probs_1_gamma = ctx.probs_1_gamma
label = ctx.label
gamma = ctx.gamma
term1 = (1.0 - probs - gamma * probs * log_probs).mul_(probs_1_gamma
).neg_()
term2 = (probs - gamma * (1.0 - probs) * log_1_probs).mul_(probs_gamma)
grads = torch.where(label == 1, term1, term2).mul_(coeff).mul_(
grad_output)
return grads, None, None, None
class Model(nn.Module):
"""
This use better formula to compute the gradient, which has better numeric stability
"""
def __init__(self, alpha=1, gamma=2, reduction='mean'):
super().__init__()
self.alpha = alpha
self.gamma = gamma
self.reduction = reduction
def forward(self, logits, label):
loss = FocalSigmoidLossFunc.apply(logits, label, self.alpha, self.gamma
)
if self.reduction == 'mean':
loss = loss.mean()
if self.reduction == 'sum':
loss = loss.sum()
return loss
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return []
|
LogisticRegression | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/gv/cgvpra4kwn5idcnpg33dwbcypnrjm2z2np7rzbwqwz3kurrqhisn.py
# Topologically Sorted Source Nodes: [mul, scores], Original ATen: [aten.mul, aten.add]
# Source node to ATen node mapping:
# mul => mul
# scores => add
# Graph fragment:
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%primals_1, %primals_2), kwargs = {})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul, %primals_3), kwargs = {})
triton_poi_fused_add_mul_0 = async_compile.triton('triton_poi_fused_add_mul_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_mul_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 3, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_mul_0(in_ptr0, in_ptr1, in_ptr2, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + (0))
tmp1 = tl.broadcast_to(tmp0, [XBLOCK])
tmp2 = tl.load(in_ptr1 + (x0), xmask)
tmp4 = tl.load(in_ptr2 + (0))
tmp5 = tl.broadcast_to(tmp4, [XBLOCK])
tmp3 = tmp1 * tmp2
tmp6 = tmp3 + tmp5
tl.store(out_ptr0 + (x0), tmp6, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (), ())
assert_size_stride(primals_2, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_3, (), ())
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [mul, scores], Original ATen: [aten.mul, aten.add]
stream0 = get_raw_stream(0)
triton_poi_fused_add_mul_0.run(primals_1, primals_2, primals_3, buf0, 256, grid=grid(256), stream=stream0)
del primals_1
del primals_3
return (buf0, primals_2, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((), (), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((), (), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
class LogisticRegression(nn.Module):
"""
A logistic regression model of the form
P(y = 1 | x) = 1 / (1 + exp(-(mx + b)))
"""
def __init__(self, init_m=1.0, init_b=1.0):
"""
Initialize a logistic regression model by defining its initial
parameters.
The ``nn.Parameter`` wrapper is needed if we intend to use vanilla
``torch.Tensors`` as parameters of this module, so that this object
knows it has trainable parameters (e.g. when calling the
``.parameters()`` method on this object, or moving the model to GPU via
``.cuda``).
If we assign parameters which are already subclasses of ``nn.Module``
(e.g. the ``nn.Linear`` layer), ``nn.Parameter`` is not needed.
Parameters
----------
init_m : ``float``, optional (defualt: 1.0)
Initial estimate for ``m``, the slope of the model
init_b : ``float``, optional (default: 1.0)
Initial estimate for ``b``, the y-intercept of the model
"""
super(LogisticRegression, self).__init__()
self.m = nn.Parameter(torch.tensor(init_m))
self.b = nn.Parameter(torch.tensor(init_b))
def forward(self, x):
"""
Forward pass through the model, which produces a prediction
``\\hat{y}`` for each given ``x`` in the batch.
Parameters
----------
x : ``torch.Tensor`` of shape ``(batch_size, )``
The single input.
Returns
-------
scores : ``torch.Tensor`` of shape ``(batch_size, )``
The raw logits of the predicted output
"""
scores = self.m * x + self.b
return scores
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {}]
| import torch
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_poi_fused_add_mul_0(in_ptr0, in_ptr1, in_ptr2, out_ptr0, xnumel,
XBLOCK: tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex
tmp0 = tl.load(in_ptr0 + 0)
tmp1 = tl.broadcast_to(tmp0, [XBLOCK])
tmp2 = tl.load(in_ptr1 + x0, xmask)
tmp4 = tl.load(in_ptr2 + 0)
tmp5 = tl.broadcast_to(tmp4, [XBLOCK])
tmp3 = tmp1 * tmp2
tmp6 = tmp3 + tmp5
tl.store(out_ptr0 + x0, tmp6, xmask)
def call(args):
primals_1, primals_2, primals_3 = args
args.clear()
assert_size_stride(primals_1, (), ())
assert_size_stride(primals_2, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_3, (), ())
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_add_mul_0[grid(256)](primals_1, primals_2,
primals_3, buf0, 256, XBLOCK=128, num_warps=4, num_stages=1)
del primals_1
del primals_3
return buf0, primals_2
class LogisticRegressionNew(nn.Module):
"""
A logistic regression model of the form
P(y = 1 | x) = 1 / (1 + exp(-(mx + b)))
"""
def __init__(self, init_m=1.0, init_b=1.0):
"""
Initialize a logistic regression model by defining its initial
parameters.
The ``nn.Parameter`` wrapper is needed if we intend to use vanilla
``torch.Tensors`` as parameters of this module, so that this object
knows it has trainable parameters (e.g. when calling the
``.parameters()`` method on this object, or moving the model to GPU via
``.cuda``).
If we assign parameters which are already subclasses of ``nn.Module``
(e.g. the ``nn.Linear`` layer), ``nn.Parameter`` is not needed.
Parameters
----------
init_m : ``float``, optional (defualt: 1.0)
Initial estimate for ``m``, the slope of the model
init_b : ``float``, optional (default: 1.0)
Initial estimate for ``b``, the y-intercept of the model
"""
super(LogisticRegressionNew, self).__init__()
self.m = nn.Parameter(torch.tensor(init_m))
self.b = nn.Parameter(torch.tensor(init_b))
def forward(self, input_0):
primals_1 = self.m
primals_3 = self.b
primals_2 = input_0
output = call([primals_1, primals_2, primals_3])
return output[0]
| jayelm/pytorch-project-template | LogisticRegression | false | 6,924 | [
"MIT"
] | 1 | 30306ce07b21c97c6993432764cbbe0a73092a0c | https://github.com/jayelm/pytorch-project-template/tree/30306ce07b21c97c6993432764cbbe0a73092a0c | import torch
import torch.nn as nn
class Model(nn.Module):
"""
A logistic regression model of the form
P(y = 1 | x) = 1 / (1 + exp(-(mx + b)))
"""
def __init__(self, init_m=1.0, init_b=1.0):
"""
Initialize a logistic regression model by defining its initial
parameters.
The ``nn.Parameter`` wrapper is needed if we intend to use vanilla
``torch.Tensors`` as parameters of this module, so that this object
knows it has trainable parameters (e.g. when calling the
``.parameters()`` method on this object, or moving the model to GPU via
``.cuda``).
If we assign parameters which are already subclasses of ``nn.Module``
(e.g. the ``nn.Linear`` layer), ``nn.Parameter`` is not needed.
Parameters
----------
init_m : ``float``, optional (defualt: 1.0)
Initial estimate for ``m``, the slope of the model
init_b : ``float``, optional (default: 1.0)
Initial estimate for ``b``, the y-intercept of the model
"""
super().__init__()
self.m = nn.Parameter(torch.tensor(init_m))
self.b = nn.Parameter(torch.tensor(init_b))
def forward(self, x):
"""
Forward pass through the model, which produces a prediction
``\\hat{y}`` for each given ``x`` in the batch.
Parameters
----------
x : ``torch.Tensor`` of shape ``(batch_size, )``
The single input.
Returns
-------
scores : ``torch.Tensor`` of shape ``(batch_size, )``
The raw logits of the predicted output
"""
scores = self.m * x + self.b
return scores
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return []
|
DownConvBlock | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/tw/ctwvjh3sk72wvpvfe6qttbbnuv7go7omfvhyfpoli3k62h5jasad.py
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.max_pool2d_with_indices]
# Source node to ATen node mapping:
# x => getitem
# Graph fragment:
# %getitem : [num_users=2] = call_function[target=operator.getitem](args = (%_low_memory_max_pool2d_with_offsets, 0), kwargs = {})
triton_poi_fused_max_pool2d_with_indices_0 = async_compile.triton('triton_poi_fused_max_pool2d_with_indices_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_max_pool2d_with_indices_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 4, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_max_pool2d_with_indices_0(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 2
x1 = (xindex // 2)
x2 = xindex
tmp0 = tl.load(in_ptr0 + ((2*x0) + (8*x1)), xmask, eviction_policy='evict_last')
tmp1 = tl.load(in_ptr0 + (1 + (2*x0) + (8*x1)), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr0 + (4 + (2*x0) + (8*x1)), xmask, eviction_policy='evict_last')
tmp5 = tl.load(in_ptr0 + (5 + (2*x0) + (8*x1)), xmask, eviction_policy='evict_last')
tmp2 = triton_helpers.maximum(tmp1, tmp0)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp6 = triton_helpers.maximum(tmp5, tmp4)
tl.store(out_ptr0 + (x2), tmp6, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/au/cau4pihcaptiev5y2ewn2o2nvrwhk7hogc72cofmmtbyv4rxc2oy.py
# Topologically Sorted Source Nodes: [conv2d], Original ATen: [aten.convolution]
# Source node to ATen node mapping:
# conv2d => convolution
# Graph fragment:
# %convolution : [num_users=3] = call_function[target=torch.ops.aten.convolution.default](args = (%getitem, %primals_2, %primals_3, [1, 1], [1, 1], [1, 1], False, [0, 0], 1), kwargs = {})
triton_poi_fused_convolution_1 = async_compile.triton('triton_poi_fused_convolution_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_convolution_1', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_convolution_1(in_out_ptr0, in_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = (xindex // 4) % 4
tmp0 = tl.load(in_out_ptr0 + (x3), xmask)
tmp1 = tl.load(in_ptr0 + (x1), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + (x3), tmp2, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/um/cumaohg52pofzqvgg5qjxgmci4axxuo6g3fgsw4oge7f3s4szluc.py
# Topologically Sorted Source Nodes: [batch_norm], Original ATen: [aten._native_batch_norm_legit]
# Source node to ATen node mapping:
# batch_norm => add, rsqrt, var_mean
# Graph fragment:
# %var_mean : [num_users=2] = call_function[target=torch.ops.aten.var_mean.correction](args = (%convolution, [0, 2, 3]), kwargs = {correction: 0, keepdim: True})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%getitem_2, 1e-05), kwargs = {})
# %rsqrt : [num_users=2] = call_function[target=torch.ops.aten.rsqrt.default](args = (%add,), kwargs = {})
triton_per_fused__native_batch_norm_legit_2 = async_compile.triton('triton_per_fused__native_batch_norm_legit_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[4, 16],
reduction_hint=ReductionHint.DEFAULT,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32', 5: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 5), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused__native_batch_norm_legit_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 4, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused__native_batch_norm_legit_2(in_ptr0, out_ptr0, out_ptr1, out_ptr2, xnumel, rnumel, XBLOCK : tl.constexpr):
xnumel = 4
rnumel = 16
RBLOCK: tl.constexpr = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
roffset = 0
rmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
r1 = rindex % 4
r2 = (rindex // 4)
x0 = xindex
tmp0 = tl.load(in_ptr0 + (r1 + (4*x0) + (16*r2)), xmask, other=0.0)
tmp1 = tl.broadcast_to(tmp0, [XBLOCK, RBLOCK])
tmp3 = tl.where(xmask, tmp1, 0)
tmp4 = tl.broadcast_to(tmp1, [XBLOCK, RBLOCK])
tmp6 = tl.where(xmask, tmp4, 0)
tmp7 = tl.sum(tmp6, 1)[:, None]
tmp8 = tl.full([XBLOCK, 1], 16, tl.int32)
tmp9 = tmp8.to(tl.float32)
tmp10 = tmp7 / tmp9
tmp11 = tmp1 - tmp10
tmp12 = tmp11 * tmp11
tmp13 = tl.broadcast_to(tmp12, [XBLOCK, RBLOCK])
tmp15 = tl.where(xmask, tmp13, 0)
tmp16 = tl.sum(tmp15, 1)[:, None]
tmp17 = 16.0
tmp18 = tmp16 / tmp17
tmp19 = 1e-05
tmp20 = tmp18 + tmp19
tmp21 = libdevice.rsqrt(tmp20)
tl.store(out_ptr2 + (x0), tmp21, xmask)
tl.store(out_ptr0 + (x0), tmp10, xmask)
tl.store(out_ptr1 + (x0), tmp16, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/hn/chnmfpowwgixhsz4q4awqox3ykrt6qsx7vjhrglkfzqjlsscicim.py
# Topologically Sorted Source Nodes: [batch_norm, x_1], Original ATen: [aten._native_batch_norm_legit, aten.relu]
# Source node to ATen node mapping:
# batch_norm => add, add_1, mul, mul_1, rsqrt, sub, var_mean
# x_1 => relu
# Graph fragment:
# %var_mean : [num_users=2] = call_function[target=torch.ops.aten.var_mean.correction](args = (%convolution, [0, 2, 3]), kwargs = {correction: 0, keepdim: True})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%getitem_2, 1e-05), kwargs = {})
# %rsqrt : [num_users=2] = call_function[target=torch.ops.aten.rsqrt.default](args = (%add,), kwargs = {})
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%convolution, %getitem_3), kwargs = {})
# %mul : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sub, %rsqrt), kwargs = {})
# %mul_1 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%mul, %unsqueeze_1), kwargs = {})
# %add_1 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul_1, %unsqueeze_3), kwargs = {})
# %relu : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%add_1,), kwargs = {})
triton_poi_fused__native_batch_norm_legit_relu_3 = async_compile.triton('triton_poi_fused__native_batch_norm_legit_relu_3', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: '*fp32', 5: '*fp32', 6: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5, 6), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__native_batch_norm_legit_relu_3', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__native_batch_norm_legit_relu_3(in_ptr0, in_ptr1, in_ptr2, in_ptr3, in_ptr4, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = (xindex // 4) % 4
tmp0 = tl.load(in_ptr0 + (x3), xmask)
tmp1 = tl.load(in_ptr1 + (x1), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr2 + (x1), xmask, eviction_policy='evict_last')
tmp10 = tl.load(in_ptr3 + (x1), xmask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr4 + (x1), xmask, eviction_policy='evict_last')
tmp2 = tmp0 - tmp1
tmp4 = 16.0
tmp5 = tmp3 / tmp4
tmp6 = 1e-05
tmp7 = tmp5 + tmp6
tmp8 = libdevice.rsqrt(tmp7)
tmp9 = tmp2 * tmp8
tmp11 = tmp9 * tmp10
tmp13 = tmp11 + tmp12
tmp14 = tl.full([1], 0, tl.int32)
tmp15 = triton_helpers.maximum(tmp14, tmp13)
tl.store(out_ptr0 + (x3), tmp15, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/bv/cbv7udtpzuwoyppfybutfnddsxgutbtpilr3fgojhdronikl2zgl.py
# Topologically Sorted Source Nodes: [batch_norm_2, x_3], Original ATen: [aten._native_batch_norm_legit, aten.relu, aten.threshold_backward]
# Source node to ATen node mapping:
# batch_norm_2 => add_4, add_5, mul_4, mul_5, rsqrt_2, sub_2, var_mean_2
# x_3 => relu_2
# Graph fragment:
# %var_mean_2 : [num_users=2] = call_function[target=torch.ops.aten.var_mean.correction](args = (%convolution_2, [0, 2, 3]), kwargs = {correction: 0, keepdim: True})
# %add_4 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%getitem_6, 1e-05), kwargs = {})
# %rsqrt_2 : [num_users=2] = call_function[target=torch.ops.aten.rsqrt.default](args = (%add_4,), kwargs = {})
# %sub_2 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%convolution_2, %getitem_7), kwargs = {})
# %mul_4 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%sub_2, %rsqrt_2), kwargs = {})
# %mul_5 : [num_users=1] = call_function[target=torch.ops.aten.mul.Tensor](args = (%mul_4, %unsqueeze_1), kwargs = {})
# %add_5 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mul_5, %unsqueeze_3), kwargs = {})
# %relu_2 : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%add_5,), kwargs = {})
# %le : [num_users=1] = call_function[target=torch.ops.aten.le.Scalar](args = (%relu_2, 0), kwargs = {})
triton_poi_fused__native_batch_norm_legit_relu_threshold_backward_4 = async_compile.triton('triton_poi_fused__native_batch_norm_legit_relu_threshold_backward_4', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: '*fp32', 5: '*fp32', 6: '*i1', 7: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5, 6, 7), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__native_batch_norm_legit_relu_threshold_backward_4', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__native_batch_norm_legit_relu_threshold_backward_4(in_ptr0, in_ptr1, in_ptr2, in_ptr3, in_ptr4, out_ptr0, out_ptr1, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = (xindex // 4) % 4
tmp0 = tl.load(in_ptr0 + (x3), xmask)
tmp1 = tl.load(in_ptr1 + (x1), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr2 + (x1), xmask, eviction_policy='evict_last')
tmp10 = tl.load(in_ptr3 + (x1), xmask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr4 + (x1), xmask, eviction_policy='evict_last')
tmp2 = tmp0 - tmp1
tmp4 = 16.0
tmp5 = tmp3 / tmp4
tmp6 = 1e-05
tmp7 = tmp5 + tmp6
tmp8 = libdevice.rsqrt(tmp7)
tmp9 = tmp2 * tmp8
tmp11 = tmp9 * tmp10
tmp13 = tmp11 + tmp12
tmp14 = tl.full([1], 0, tl.int32)
tmp15 = triton_helpers.maximum(tmp14, tmp13)
tmp16 = 0.0
tmp17 = tmp15 <= tmp16
tl.store(out_ptr0 + (x3), tmp15, xmask)
tl.store(out_ptr1 + (x3), tmp17, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9 = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (4, 4, 3, 3), (36, 9, 3, 1))
assert_size_stride(primals_3, (4, ), (1, ))
assert_size_stride(primals_4, (4, ), (1, ))
assert_size_stride(primals_5, (4, ), (1, ))
assert_size_stride(primals_6, (4, 4, 3, 3), (36, 9, 3, 1))
assert_size_stride(primals_7, (4, ), (1, ))
assert_size_stride(primals_8, (4, 4, 3, 3), (36, 9, 3, 1))
assert_size_stride(primals_9, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 2, 2), (16, 4, 2, 1), torch.float32)
# Topologically Sorted Source Nodes: [x], Original ATen: [aten.max_pool2d_with_indices]
stream0 = get_raw_stream(0)
triton_poi_fused_max_pool2d_with_indices_0.run(primals_1, buf0, 64, grid=grid(64), stream=stream0)
del primals_1
# Topologically Sorted Source Nodes: [conv2d], Original ATen: [aten.convolution]
buf1 = extern_kernels.convolution(buf0, primals_2, stride=(1, 1), padding=(1, 1), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf1, (4, 4, 2, 2), (16, 4, 2, 1))
buf2 = buf1; del buf1 # reuse
# Topologically Sorted Source Nodes: [conv2d], Original ATen: [aten.convolution]
triton_poi_fused_convolution_1.run(buf2, primals_3, 64, grid=grid(64), stream=stream0)
del primals_3
buf3 = empty_strided_cuda((1, 4, 1, 1), (4, 1, 4, 4), torch.float32)
buf4 = empty_strided_cuda((1, 4, 1, 1), (4, 1, 4, 4), torch.float32)
buf6 = empty_strided_cuda((1, 4, 1, 1), (4, 1, 4, 4), torch.float32)
# Topologically Sorted Source Nodes: [batch_norm], Original ATen: [aten._native_batch_norm_legit]
triton_per_fused__native_batch_norm_legit_2.run(buf2, buf3, buf4, buf6, 4, 16, grid=grid(4), stream=stream0)
buf7 = empty_strided_cuda((4, 4, 2, 2), (16, 4, 2, 1), torch.float32)
# Topologically Sorted Source Nodes: [batch_norm, x_1], Original ATen: [aten._native_batch_norm_legit, aten.relu]
triton_poi_fused__native_batch_norm_legit_relu_3.run(buf2, buf3, buf4, primals_4, primals_5, buf7, 64, grid=grid(64), stream=stream0)
# Topologically Sorted Source Nodes: [conv2d_1], Original ATen: [aten.convolution]
buf8 = extern_kernels.convolution(buf7, primals_6, stride=(1, 1), padding=(1, 1), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf8, (4, 4, 2, 2), (16, 4, 2, 1))
buf9 = buf8; del buf8 # reuse
# Topologically Sorted Source Nodes: [conv2d_1], Original ATen: [aten.convolution]
triton_poi_fused_convolution_1.run(buf9, primals_7, 64, grid=grid(64), stream=stream0)
del primals_7
buf10 = buf4; del buf4 # reuse
buf11 = empty_strided_cuda((1, 4, 1, 1), (4, 1, 4, 4), torch.float32)
buf13 = empty_strided_cuda((1, 4, 1, 1), (4, 1, 4, 4), torch.float32)
# Topologically Sorted Source Nodes: [batch_norm_1], Original ATen: [aten._native_batch_norm_legit]
triton_per_fused__native_batch_norm_legit_2.run(buf9, buf10, buf11, buf13, 4, 16, grid=grid(4), stream=stream0)
buf14 = empty_strided_cuda((4, 4, 2, 2), (16, 4, 2, 1), torch.float32)
# Topologically Sorted Source Nodes: [batch_norm_1, x_2], Original ATen: [aten._native_batch_norm_legit, aten.relu]
triton_poi_fused__native_batch_norm_legit_relu_3.run(buf9, buf10, buf11, primals_4, primals_5, buf14, 64, grid=grid(64), stream=stream0)
# Topologically Sorted Source Nodes: [conv2d_2], Original ATen: [aten.convolution]
buf15 = extern_kernels.convolution(buf14, primals_8, stride=(1, 1), padding=(1, 1), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf15, (4, 4, 2, 2), (16, 4, 2, 1))
buf16 = buf15; del buf15 # reuse
# Topologically Sorted Source Nodes: [conv2d_2], Original ATen: [aten.convolution]
triton_poi_fused_convolution_1.run(buf16, primals_9, 64, grid=grid(64), stream=stream0)
del primals_9
buf17 = buf11; del buf11 # reuse
buf18 = empty_strided_cuda((1, 4, 1, 1), (4, 1, 4, 4), torch.float32)
buf20 = empty_strided_cuda((1, 4, 1, 1), (4, 1, 4, 4), torch.float32)
# Topologically Sorted Source Nodes: [batch_norm_2], Original ATen: [aten._native_batch_norm_legit]
triton_per_fused__native_batch_norm_legit_2.run(buf16, buf17, buf18, buf20, 4, 16, grid=grid(4), stream=stream0)
buf21 = empty_strided_cuda((4, 4, 2, 2), (16, 4, 2, 1), torch.float32)
buf22 = empty_strided_cuda((4, 4, 2, 2), (16, 4, 2, 1), torch.bool)
# Topologically Sorted Source Nodes: [batch_norm_2, x_3], Original ATen: [aten._native_batch_norm_legit, aten.relu, aten.threshold_backward]
triton_poi_fused__native_batch_norm_legit_relu_threshold_backward_4.run(buf16, buf17, buf18, primals_4, primals_5, buf21, buf22, 64, grid=grid(64), stream=stream0)
del buf18
del primals_5
return (buf21, primals_2, primals_4, primals_6, primals_8, buf0, buf2, reinterpret_tensor(buf6, (4, ), (1, ), 0), buf7, buf9, reinterpret_tensor(buf13, (4, ), (1, ), 0), buf14, buf16, reinterpret_tensor(buf20, (4, ), (1, ), 0), buf22, reinterpret_tensor(buf17, (1, 4, 1, 1), (4, 1, 1, 1), 0), reinterpret_tensor(buf10, (1, 4, 1, 1), (4, 1, 1, 1), 0), reinterpret_tensor(buf3, (1, 4, 1, 1), (4, 1, 1, 1), 0), )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((4, 4, 3, 3), (36, 9, 3, 1), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((4, 4, 3, 3), (36, 9, 3, 1), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_8 = rand_strided((4, 4, 3, 3), (36, 9, 3, 1), device='cuda:0', dtype=torch.float32)
primals_9 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7, primals_8, primals_9])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.utils.data
from torch.cuda import *
def conv3x3(in_channels, out_channels):
return nn.Conv2d(in_channels, out_channels, kernel_size=3, stride=1,
padding=1, bias=True)
def maxpool2x2():
return nn.MaxPool2d(kernel_size=2, stride=2, padding=0)
class DownConvBlock(nn.Module):
def __init__(self, in_channels, out_channels):
"""
Args:
in_channels: number of channels in input (1st) feature map
out_channels: number of channels in output feature maps
"""
super(DownConvBlock, self).__init__()
self.maxpool = maxpool2x2()
self.conv1 = conv3x3(in_channels, out_channels)
self.conv2 = conv3x3(out_channels, out_channels)
self.conv3 = conv3x3(out_channels, out_channels)
self.norm = nn.BatchNorm2d(out_channels, track_running_stats=False)
def forward(self, x):
x = self.maxpool(x)
x = F.relu(self.norm(self.conv1(x)))
x = F.relu(self.norm(self.conv2(x)))
x = F.relu(self.norm(self.conv3(x)))
return x
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'in_channels': 4, 'out_channels': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import libdevice
import torch.nn as nn
import torch.utils.data
from torch.cuda import *
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_max_pool2d_with_indices_0(in_ptr0, out_ptr0, xnumel,
XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 2
x1 = xindex // 2
x2 = xindex
tmp0 = tl.load(in_ptr0 + (2 * x0 + 8 * x1), xmask, eviction_policy=
'evict_last')
tmp1 = tl.load(in_ptr0 + (1 + 2 * x0 + 8 * x1), xmask, eviction_policy=
'evict_last')
tmp3 = tl.load(in_ptr0 + (4 + 2 * x0 + 8 * x1), xmask, eviction_policy=
'evict_last')
tmp5 = tl.load(in_ptr0 + (5 + 2 * x0 + 8 * x1), xmask, eviction_policy=
'evict_last')
tmp2 = triton_helpers.maximum(tmp1, tmp0)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp6 = triton_helpers.maximum(tmp5, tmp4)
tl.store(out_ptr0 + x2, tmp6, xmask)
@triton.jit
def triton_poi_fused_convolution_1(in_out_ptr0, in_ptr0, xnumel, XBLOCK: tl
.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = xindex // 4 % 4
tmp0 = tl.load(in_out_ptr0 + x3, xmask)
tmp1 = tl.load(in_ptr0 + x1, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tl.store(in_out_ptr0 + x3, tmp2, xmask)
@triton.jit
def triton_per_fused__native_batch_norm_legit_2(in_ptr0, out_ptr0, out_ptr1,
out_ptr2, xnumel, rnumel, XBLOCK: tl.constexpr):
xnumel = 4
RBLOCK: tl.constexpr = 16
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
rindex = tl.arange(0, RBLOCK)[None, :]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
r1 = rindex % 4
r2 = rindex // 4
x0 = xindex
tmp0 = tl.load(in_ptr0 + (r1 + 4 * x0 + 16 * r2), xmask, other=0.0)
tmp1 = tl.broadcast_to(tmp0, [XBLOCK, RBLOCK])
tl.where(xmask, tmp1, 0)
tmp4 = tl.broadcast_to(tmp1, [XBLOCK, RBLOCK])
tmp6 = tl.where(xmask, tmp4, 0)
tmp7 = tl.sum(tmp6, 1)[:, None]
tmp8 = tl.full([XBLOCK, 1], 16, tl.int32)
tmp9 = tmp8.to(tl.float32)
tmp10 = tmp7 / tmp9
tmp11 = tmp1 - tmp10
tmp12 = tmp11 * tmp11
tmp13 = tl.broadcast_to(tmp12, [XBLOCK, RBLOCK])
tmp15 = tl.where(xmask, tmp13, 0)
tmp16 = tl.sum(tmp15, 1)[:, None]
tmp17 = 16.0
tmp18 = tmp16 / tmp17
tmp19 = 1e-05
tmp20 = tmp18 + tmp19
tmp21 = libdevice.rsqrt(tmp20)
tl.store(out_ptr2 + x0, tmp21, xmask)
tl.store(out_ptr0 + x0, tmp10, xmask)
tl.store(out_ptr1 + x0, tmp16, xmask)
@triton.jit
def triton_poi_fused__native_batch_norm_legit_relu_3(in_ptr0, in_ptr1,
in_ptr2, in_ptr3, in_ptr4, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = xindex // 4 % 4
tmp0 = tl.load(in_ptr0 + x3, xmask)
tmp1 = tl.load(in_ptr1 + x1, xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr2 + x1, xmask, eviction_policy='evict_last')
tmp10 = tl.load(in_ptr3 + x1, xmask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr4 + x1, xmask, eviction_policy='evict_last')
tmp2 = tmp0 - tmp1
tmp4 = 16.0
tmp5 = tmp3 / tmp4
tmp6 = 1e-05
tmp7 = tmp5 + tmp6
tmp8 = libdevice.rsqrt(tmp7)
tmp9 = tmp2 * tmp8
tmp11 = tmp9 * tmp10
tmp13 = tmp11 + tmp12
tmp14 = tl.full([1], 0, tl.int32)
tmp15 = triton_helpers.maximum(tmp14, tmp13)
tl.store(out_ptr0 + x3, tmp15, xmask)
@triton.jit
def triton_poi_fused__native_batch_norm_legit_relu_threshold_backward_4(in_ptr0
, in_ptr1, in_ptr2, in_ptr3, in_ptr4, out_ptr0, out_ptr1, xnumel,
XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x1 = xindex // 4 % 4
tmp0 = tl.load(in_ptr0 + x3, xmask)
tmp1 = tl.load(in_ptr1 + x1, xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr2 + x1, xmask, eviction_policy='evict_last')
tmp10 = tl.load(in_ptr3 + x1, xmask, eviction_policy='evict_last')
tmp12 = tl.load(in_ptr4 + x1, xmask, eviction_policy='evict_last')
tmp2 = tmp0 - tmp1
tmp4 = 16.0
tmp5 = tmp3 / tmp4
tmp6 = 1e-05
tmp7 = tmp5 + tmp6
tmp8 = libdevice.rsqrt(tmp7)
tmp9 = tmp2 * tmp8
tmp11 = tmp9 * tmp10
tmp13 = tmp11 + tmp12
tmp14 = tl.full([1], 0, tl.int32)
tmp15 = triton_helpers.maximum(tmp14, tmp13)
tmp16 = 0.0
tmp17 = tmp15 <= tmp16
tl.store(out_ptr0 + x3, tmp15, xmask)
tl.store(out_ptr1 + x3, tmp17, xmask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7, primals_8, primals_9) = args
args.clear()
assert_size_stride(primals_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_2, (4, 4, 3, 3), (36, 9, 3, 1))
assert_size_stride(primals_3, (4,), (1,))
assert_size_stride(primals_4, (4,), (1,))
assert_size_stride(primals_5, (4,), (1,))
assert_size_stride(primals_6, (4, 4, 3, 3), (36, 9, 3, 1))
assert_size_stride(primals_7, (4,), (1,))
assert_size_stride(primals_8, (4, 4, 3, 3), (36, 9, 3, 1))
assert_size_stride(primals_9, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 2, 2), (16, 4, 2, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_max_pool2d_with_indices_0[grid(64)](primals_1,
buf0, 64, XBLOCK=64, num_warps=1, num_stages=1)
del primals_1
buf1 = extern_kernels.convolution(buf0, primals_2, stride=(1, 1),
padding=(1, 1), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf1, (4, 4, 2, 2), (16, 4, 2, 1))
buf2 = buf1
del buf1
triton_poi_fused_convolution_1[grid(64)](buf2, primals_3, 64,
XBLOCK=64, num_warps=1, num_stages=1)
del primals_3
buf3 = empty_strided_cuda((1, 4, 1, 1), (4, 1, 4, 4), torch.float32)
buf4 = empty_strided_cuda((1, 4, 1, 1), (4, 1, 4, 4), torch.float32)
buf6 = empty_strided_cuda((1, 4, 1, 1), (4, 1, 4, 4), torch.float32)
triton_per_fused__native_batch_norm_legit_2[grid(4)](buf2, buf3,
buf4, buf6, 4, 16, XBLOCK=1, num_warps=2, num_stages=1)
buf7 = empty_strided_cuda((4, 4, 2, 2), (16, 4, 2, 1), torch.float32)
triton_poi_fused__native_batch_norm_legit_relu_3[grid(64)](buf2,
buf3, buf4, primals_4, primals_5, buf7, 64, XBLOCK=64,
num_warps=1, num_stages=1)
buf8 = extern_kernels.convolution(buf7, primals_6, stride=(1, 1),
padding=(1, 1), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf8, (4, 4, 2, 2), (16, 4, 2, 1))
buf9 = buf8
del buf8
triton_poi_fused_convolution_1[grid(64)](buf9, primals_7, 64,
XBLOCK=64, num_warps=1, num_stages=1)
del primals_7
buf10 = buf4
del buf4
buf11 = empty_strided_cuda((1, 4, 1, 1), (4, 1, 4, 4), torch.float32)
buf13 = empty_strided_cuda((1, 4, 1, 1), (4, 1, 4, 4), torch.float32)
triton_per_fused__native_batch_norm_legit_2[grid(4)](buf9, buf10,
buf11, buf13, 4, 16, XBLOCK=1, num_warps=2, num_stages=1)
buf14 = empty_strided_cuda((4, 4, 2, 2), (16, 4, 2, 1), torch.float32)
triton_poi_fused__native_batch_norm_legit_relu_3[grid(64)](buf9,
buf10, buf11, primals_4, primals_5, buf14, 64, XBLOCK=64,
num_warps=1, num_stages=1)
buf15 = extern_kernels.convolution(buf14, primals_8, stride=(1, 1),
padding=(1, 1), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf15, (4, 4, 2, 2), (16, 4, 2, 1))
buf16 = buf15
del buf15
triton_poi_fused_convolution_1[grid(64)](buf16, primals_9, 64,
XBLOCK=64, num_warps=1, num_stages=1)
del primals_9
buf17 = buf11
del buf11
buf18 = empty_strided_cuda((1, 4, 1, 1), (4, 1, 4, 4), torch.float32)
buf20 = empty_strided_cuda((1, 4, 1, 1), (4, 1, 4, 4), torch.float32)
triton_per_fused__native_batch_norm_legit_2[grid(4)](buf16, buf17,
buf18, buf20, 4, 16, XBLOCK=1, num_warps=2, num_stages=1)
buf21 = empty_strided_cuda((4, 4, 2, 2), (16, 4, 2, 1), torch.float32)
buf22 = empty_strided_cuda((4, 4, 2, 2), (16, 4, 2, 1), torch.bool)
triton_poi_fused__native_batch_norm_legit_relu_threshold_backward_4[
grid(64)](buf16, buf17, buf18, primals_4, primals_5, buf21,
buf22, 64, XBLOCK=64, num_warps=1, num_stages=1)
del buf18
del primals_5
return (buf21, primals_2, primals_4, primals_6, primals_8, buf0, buf2,
reinterpret_tensor(buf6, (4,), (1,), 0), buf7, buf9,
reinterpret_tensor(buf13, (4,), (1,), 0), buf14, buf16,
reinterpret_tensor(buf20, (4,), (1,), 0), buf22, reinterpret_tensor
(buf17, (1, 4, 1, 1), (4, 1, 1, 1), 0), reinterpret_tensor(buf10, (
1, 4, 1, 1), (4, 1, 1, 1), 0), reinterpret_tensor(buf3, (1, 4, 1, 1
), (4, 1, 1, 1), 0))
def conv3x3(in_channels, out_channels):
return nn.Conv2d(in_channels, out_channels, kernel_size=3, stride=1,
padding=1, bias=True)
def maxpool2x2():
return nn.MaxPool2d(kernel_size=2, stride=2, padding=0)
class DownConvBlockNew(nn.Module):
def __init__(self, in_channels, out_channels):
"""
Args:
in_channels: number of channels in input (1st) feature map
out_channels: number of channels in output feature maps
"""
super(DownConvBlockNew, self).__init__()
self.maxpool = maxpool2x2()
self.conv1 = conv3x3(in_channels, out_channels)
self.conv2 = conv3x3(out_channels, out_channels)
self.conv3 = conv3x3(out_channels, out_channels)
self.norm = nn.BatchNorm2d(out_channels, track_running_stats=False)
def forward(self, input_0):
primals_2 = self.conv1.weight
primals_3 = self.conv1.bias
primals_6 = self.conv2.weight
primals_4 = self.conv2.bias
primals_8 = self.conv3.weight
primals_5 = self.conv3.bias
primals_7 = self.norm.weight
primals_9 = self.norm.bias
primals_1 = input_0
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7, primals_8, primals_9])
return output[0]
| jabae/detectEM | DownConvBlock | false | 6,925 | [
"MIT"
] | 1 | 2d1a5116164d0bed0a8ea767a227d05a8970a448 | https://github.com/jabae/detectEM/tree/2d1a5116164d0bed0a8ea767a227d05a8970a448 | import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.utils.data
from torch.cuda import *
def conv3x3(in_channels, out_channels):
return nn.Conv2d(in_channels, out_channels, kernel_size=3, stride=1,
padding=1, bias=True)
def maxpool2x2():
return nn.MaxPool2d(kernel_size=2, stride=2, padding=0)
class Model(nn.Module):
def __init__(self, in_channels, out_channels):
"""
Args:
in_channels: number of channels in input (1st) feature map
out_channels: number of channels in output feature maps
"""
super().__init__()
self.maxpool = maxpool2x2()
self.conv1 = conv3x3(in_channels, out_channels)
self.conv2 = conv3x3(out_channels, out_channels)
self.conv3 = conv3x3(out_channels, out_channels)
self.norm = nn.BatchNorm2d(out_channels, track_running_stats=False)
def forward(self, x):
x = self.maxpool(x)
x = F.relu(self.norm(self.conv1(x)))
x = F.relu(self.norm(self.conv2(x)))
x = F.relu(self.norm(self.conv3(x)))
return x
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4, 4]
|
VariationalLoss | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/6s/c6sjhak22qn3fdmdl7dqnt5isergitdpjbloozsyji5zy6knc3si.py
# Topologically Sorted Source Nodes: [mse_loss, mse_loss_1, add], Original ATen: [aten.mse_loss, aten.add]
# Source node to ATen node mapping:
# add => add
# mse_loss => mean, pow_1, sub
# mse_loss_1 => mean_1, pow_2, sub_1
# Graph fragment:
# %sub : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%slice_3, %slice_7), kwargs = {})
# %pow_1 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%sub, 2), kwargs = {})
# %mean : [num_users=1] = call_function[target=torch.ops.aten.mean.default](args = (%pow_1,), kwargs = {})
# %sub_1 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%slice_12, %slice_16), kwargs = {})
# %pow_2 : [num_users=1] = call_function[target=torch.ops.aten.pow.Tensor_Scalar](args = (%sub_1, 2), kwargs = {})
# %mean_1 : [num_users=1] = call_function[target=torch.ops.aten.mean.default](args = (%pow_2,), kwargs = {})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%mean, %mean_1), kwargs = {})
triton_per_fused_add_mse_loss_0 = async_compile.triton('triton_per_fused_add_mse_loss_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.persistent_reduction(
size_hints=[1, 256],
reduction_hint=ReductionHint.INNER,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {2: 1}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 3), equal_to_1=(2,))]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_per_fused_add_mse_loss_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 4, 'num_reduction': 2, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False}
)
@triton.jit
def triton_per_fused_add_mse_loss_0(in_out_ptr0, in_ptr0, xnumel, rnumel, XBLOCK : tl.constexpr):
xnumel = 1
rnumel = 192
RBLOCK: tl.constexpr = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = tl.full([XBLOCK, RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[None, :]
roffset = 0
rmask = rindex < rnumel
r0 = rindex % 12
r1 = (rindex // 12)
r2 = rindex % 3
r3 = (rindex // 3)
tmp0 = tl.load(in_ptr0 + (4 + r0 + (16*r1)), rmask, other=0.0)
tmp1 = tl.load(in_ptr0 + (r0 + (16*r1)), rmask, other=0.0)
tmp8 = tl.load(in_ptr0 + (1 + r2 + (4*r3)), rmask, other=0.0)
tmp9 = tl.load(in_ptr0 + (r2 + (4*r3)), rmask, other=0.0)
tmp2 = tmp0 - tmp1
tmp3 = tmp2 * tmp2
tmp4 = tl.broadcast_to(tmp3, [XBLOCK, RBLOCK])
tmp6 = tl.where(rmask, tmp4, 0)
tmp7 = tl.sum(tmp6, 1)[:, None]
tmp10 = tmp8 - tmp9
tmp11 = tmp10 * tmp10
tmp12 = tl.broadcast_to(tmp11, [XBLOCK, RBLOCK])
tmp14 = tl.where(rmask, tmp12, 0)
tmp15 = tl.sum(tmp14, 1)[:, None]
tmp16 = 192.0
tmp17 = tmp7 / tmp16
tmp18 = tmp15 / tmp16
tmp19 = tmp17 + tmp18
tl.debug_barrier()
tl.store(in_out_ptr0 + (tl.full([XBLOCK, 1], 0, tl.int32)), tmp19, None)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((), (), torch.float32)
buf2 = buf0; del buf0 # reuse
# Topologically Sorted Source Nodes: [mse_loss, mse_loss_1, add], Original ATen: [aten.mse_loss, aten.add]
stream0 = get_raw_stream(0)
triton_per_fused_add_mse_loss_0.run(buf2, arg0_1, 1, 192, grid=grid(1), stream=stream0)
del arg0_1
return (buf2, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([arg0_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.nn.functional as F
class VariationalLoss(nn.Module):
""" Variational loss to enforce continuity of images
"""
def forward(self, input):
""" forward pass """
self.loss = F.mse_loss(input[:, :, 1:, :], input[:, :, :-1, :]
) + F.mse_loss(input[:, :, :, 1:], input[:, :, :, :-1])
return input
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {}]
| import torch
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_per_fused_add_mse_loss_0(in_out_ptr0, in_ptr0, xnumel, rnumel,
XBLOCK: tl.constexpr):
rnumel = 192
RBLOCK: tl.constexpr = 256
xoffset = tl.program_id(0) * XBLOCK
xoffset + tl.arange(0, XBLOCK)[:, None]
tl.full([XBLOCK, RBLOCK], True, tl.int1)
rindex = tl.arange(0, RBLOCK)[None, :]
rmask = rindex < rnumel
r0 = rindex % 12
r1 = rindex // 12
r2 = rindex % 3
r3 = rindex // 3
tmp0 = tl.load(in_ptr0 + (4 + r0 + 16 * r1), rmask, other=0.0)
tmp1 = tl.load(in_ptr0 + (r0 + 16 * r1), rmask, other=0.0)
tmp8 = tl.load(in_ptr0 + (1 + r2 + 4 * r3), rmask, other=0.0)
tmp9 = tl.load(in_ptr0 + (r2 + 4 * r3), rmask, other=0.0)
tmp2 = tmp0 - tmp1
tmp3 = tmp2 * tmp2
tmp4 = tl.broadcast_to(tmp3, [XBLOCK, RBLOCK])
tmp6 = tl.where(rmask, tmp4, 0)
tmp7 = tl.sum(tmp6, 1)[:, None]
tmp10 = tmp8 - tmp9
tmp11 = tmp10 * tmp10
tmp12 = tl.broadcast_to(tmp11, [XBLOCK, RBLOCK])
tmp14 = tl.where(rmask, tmp12, 0)
tmp15 = tl.sum(tmp14, 1)[:, None]
tmp16 = 192.0
tmp17 = tmp7 / tmp16
tmp18 = tmp15 / tmp16
tmp19 = tmp17 + tmp18
tl.debug_barrier()
tl.store(in_out_ptr0 + tl.full([XBLOCK, 1], 0, tl.int32), tmp19, None)
def call(args):
arg0_1, = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((), (), torch.float32)
buf2 = buf0
del buf0
get_raw_stream(0)
triton_per_fused_add_mse_loss_0[grid(1)](buf2, arg0_1, 1, 192,
XBLOCK=1, num_warps=2, num_stages=1)
del arg0_1
return buf2,
class VariationalLossNew(nn.Module):
""" Variational loss to enforce continuity of images
"""
def forward(self, input_0):
arg0_1 = input_0
output = call([arg0_1])
return output[0]
| jaredaevans/UltrafastNST | VariationalLoss | false | 6,926 | [
"MIT"
] | 1 | 6671c6b618ce6bb4920b15f782be962e484a5423 | https://github.com/jaredaevans/UltrafastNST/tree/6671c6b618ce6bb4920b15f782be962e484a5423 | import torch
import torch.nn as nn
import torch.nn.functional as F
class Model(nn.Module):
""" Variational loss to enforce continuity of images
"""
def forward(self, input):
""" forward pass """
self.loss = F.mse_loss(input[:, :, 1:, :], input[:, :, :-1, :]
) + F.mse_loss(input[:, :, :, 1:], input[:, :, :, :-1])
return input
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return []
|
TwoHiddenLayerFc | # AOT ID: ['0_forward']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/sr/csrxdjbtbkq5mhx4lx76hdeti625uy52jalpuc5xjwghomvl635m.py
# Topologically Sorted Source Nodes: [out], Original ATen: [aten.relu, aten.threshold_backward]
# Source node to ATen node mapping:
# out => relu
# Graph fragment:
# %relu : [num_users=2] = call_function[target=torch.ops.aten.relu.default](args = (%view_1,), kwargs = {})
# %le_1 : [num_users=1] = call_function[target=torch.ops.aten.le.Scalar](args = (%relu, 0), kwargs = {})
triton_poi_fused_relu_threshold_backward_0 = async_compile.triton('triton_poi_fused_relu_threshold_backward_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[16384],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*i1', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_relu_threshold_backward_0', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 2, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_relu_threshold_backward_0(in_out_ptr0, in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 12800
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 200
tmp0 = tl.load(in_out_ptr0 + (x2), xmask)
tmp1 = tl.load(in_ptr0 + (x0), xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(in_out_ptr0 + (x2), tmp4, xmask)
tl.store(out_ptr0 + (x2), tmp6, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7 = args
args.clear()
assert_size_stride(primals_1, (200, 4), (4, 1))
assert_size_stride(primals_2, (200, ), (1, ))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_4, (200, 200), (200, 1))
assert_size_stride(primals_5, (200, ), (1, ))
assert_size_stride(primals_6, (4, 200), (200, 1))
assert_size_stride(primals_7, (4, ), (1, ))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 200), (200, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(primals_3, (64, 4), (4, 1), 0), reinterpret_tensor(primals_1, (4, 200), (1, 4), 0), out=buf0)
del primals_1
buf1 = reinterpret_tensor(buf0, (4, 4, 4, 200), (3200, 800, 200, 1), 0); del buf0 # reuse
buf6 = empty_strided_cuda((4, 4, 4, 200), (3200, 800, 200, 1), torch.bool)
# Topologically Sorted Source Nodes: [out], Original ATen: [aten.relu, aten.threshold_backward]
stream0 = get_raw_stream(0)
triton_poi_fused_relu_threshold_backward_0.run(buf1, primals_2, buf6, 12800, grid=grid(12800), stream=stream0)
del primals_2
buf2 = empty_strided_cuda((64, 200), (200, 1), torch.float32)
# Topologically Sorted Source Nodes: [], Original ATen: []
extern_kernels.mm(reinterpret_tensor(buf1, (64, 200), (200, 1), 0), reinterpret_tensor(primals_4, (200, 200), (1, 200), 0), out=buf2)
buf3 = reinterpret_tensor(buf2, (4, 4, 4, 200), (3200, 800, 200, 1), 0); del buf2 # reuse
buf5 = empty_strided_cuda((4, 4, 4, 200), (3200, 800, 200, 1), torch.bool)
# Topologically Sorted Source Nodes: [out_1], Original ATen: [aten.relu, aten.threshold_backward]
triton_poi_fused_relu_threshold_backward_0.run(buf3, primals_5, buf5, 12800, grid=grid(12800), stream=stream0)
del primals_5
buf4 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [out_2], Original ATen: [aten.addmm]
extern_kernels.addmm(primals_7, reinterpret_tensor(buf3, (64, 200), (200, 1), 0), reinterpret_tensor(primals_6, (200, 4), (1, 200), 0), alpha=1, beta=1, out=buf4)
del primals_7
return (reinterpret_tensor(buf4, (4, 4, 4, 4), (64, 16, 4, 1), 0), reinterpret_tensor(primals_3, (64, 4), (4, 1), 0), reinterpret_tensor(buf1, (64, 200), (200, 1), 0), reinterpret_tensor(buf3, (64, 200), (200, 1), 0), primals_6, buf5, primals_4, buf6, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
primals_1 = rand_strided((200, 4), (4, 1), device='cuda:0', dtype=torch.float32)
primals_2 = rand_strided((200, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_3 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
primals_4 = rand_strided((200, 200), (200, 1), device='cuda:0', dtype=torch.float32)
primals_5 = rand_strided((200, ), (1, ), device='cuda:0', dtype=torch.float32)
primals_6 = rand_strided((4, 200), (200, 1), device='cuda:0', dtype=torch.float32)
primals_7 = rand_strided((4, ), (1, ), device='cuda:0', dtype=torch.float32)
fn = lambda: call([primals_1, primals_2, primals_3, primals_4, primals_5, primals_6, primals_7])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.nn.functional as F
class TwoHiddenLayerFc(nn.Module):
def __init__(self, input_shape, out_dim):
super(TwoHiddenLayerFc, self).__init__()
self.fc1 = nn.Linear(input_shape, 200)
self.fc2 = nn.Linear(200, 200)
self.fc3 = nn.Linear(200, out_dim)
def forward(self, x):
out = F.relu(self.fc1(x))
out = F.relu(self.fc2(out))
out = self.fc3(out)
return out
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'input_shape': 4, 'out_dim': 4}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_relu_threshold_backward_0(in_out_ptr0, in_ptr0,
out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 12800
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x2 = xindex
x0 = xindex % 200
tmp0 = tl.load(in_out_ptr0 + x2, xmask)
tmp1 = tl.load(in_ptr0 + x0, xmask, eviction_policy='evict_last')
tmp2 = tmp0 + tmp1
tmp3 = tl.full([1], 0, tl.int32)
tmp4 = triton_helpers.maximum(tmp3, tmp2)
tmp5 = 0.0
tmp6 = tmp4 <= tmp5
tl.store(in_out_ptr0 + x2, tmp4, xmask)
tl.store(out_ptr0 + x2, tmp6, xmask)
def call(args):
(primals_1, primals_2, primals_3, primals_4, primals_5, primals_6,
primals_7) = args
args.clear()
assert_size_stride(primals_1, (200, 4), (4, 1))
assert_size_stride(primals_2, (200,), (1,))
assert_size_stride(primals_3, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(primals_4, (200, 200), (200, 1))
assert_size_stride(primals_5, (200,), (1,))
assert_size_stride(primals_6, (4, 200), (200, 1))
assert_size_stride(primals_7, (4,), (1,))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((64, 200), (200, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(primals_3, (64, 4), (4, 1), 0),
reinterpret_tensor(primals_1, (4, 200), (1, 4), 0), out=buf0)
del primals_1
buf1 = reinterpret_tensor(buf0, (4, 4, 4, 200), (3200, 800, 200, 1), 0)
del buf0
buf6 = empty_strided_cuda((4, 4, 4, 200), (3200, 800, 200, 1),
torch.bool)
get_raw_stream(0)
triton_poi_fused_relu_threshold_backward_0[grid(12800)](buf1,
primals_2, buf6, 12800, XBLOCK=128, num_warps=4, num_stages=1)
del primals_2
buf2 = empty_strided_cuda((64, 200), (200, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(buf1, (64, 200), (200, 1), 0),
reinterpret_tensor(primals_4, (200, 200), (1, 200), 0), out=buf2)
buf3 = reinterpret_tensor(buf2, (4, 4, 4, 200), (3200, 800, 200, 1), 0)
del buf2
buf5 = empty_strided_cuda((4, 4, 4, 200), (3200, 800, 200, 1),
torch.bool)
triton_poi_fused_relu_threshold_backward_0[grid(12800)](buf3,
primals_5, buf5, 12800, XBLOCK=128, num_warps=4, num_stages=1)
del primals_5
buf4 = empty_strided_cuda((64, 4), (4, 1), torch.float32)
extern_kernels.addmm(primals_7, reinterpret_tensor(buf3, (64, 200),
(200, 1), 0), reinterpret_tensor(primals_6, (200, 4), (1, 200),
0), alpha=1, beta=1, out=buf4)
del primals_7
return reinterpret_tensor(buf4, (4, 4, 4, 4), (64, 16, 4, 1), 0
), reinterpret_tensor(primals_3, (64, 4), (4, 1), 0
), reinterpret_tensor(buf1, (64, 200), (200, 1), 0
), reinterpret_tensor(buf3, (64, 200), (200, 1), 0
), primals_6, buf5, primals_4, buf6
class TwoHiddenLayerFcNew(nn.Module):
def __init__(self, input_shape, out_dim):
super(TwoHiddenLayerFcNew, self).__init__()
self.fc1 = nn.Linear(input_shape, 200)
self.fc2 = nn.Linear(200, 200)
self.fc3 = nn.Linear(200, out_dim)
def forward(self, input_0):
primals_1 = self.fc1.weight
primals_2 = self.fc1.bias
primals_4 = self.fc2.weight
primals_5 = self.fc2.bias
primals_6 = self.fc3.weight
primals_7 = self.fc3.bias
primals_3 = input_0
output = call([primals_1, primals_2, primals_3, primals_4,
primals_5, primals_6, primals_7])
return output[0]
| jasonyanglu/fedavgpy | TwoHiddenLayerFc | false | 6,927 | [
"MIT"
] | 1 | cefbe5854f02d3df1197d849872286439c86e949 | https://github.com/jasonyanglu/fedavgpy/tree/cefbe5854f02d3df1197d849872286439c86e949 | import torch
import torch.nn as nn
import torch.nn.functional as F
class Model(nn.Module):
def __init__(self, input_shape, out_dim):
super().__init__()
self.fc1 = nn.Linear(input_shape, 200)
self.fc2 = nn.Linear(200, 200)
self.fc3 = nn.Linear(200, out_dim)
def forward(self, x):
out = F.relu(self.fc1(x))
out = F.relu(self.fc2(out))
out = self.fc3(out)
return out
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4, 4]
|
SoftCrossEntropyLoss2d | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/td/ctdj5kazgiki6gdaadhqtp2x7tq2ee5ey5hqqdcoqmp54jyhf74f.py
# Topologically Sorted Source Nodes: [log_softmax], Original ATen: [aten._log_softmax]
# Source node to ATen node mapping:
# log_softmax => amax, sub
# Graph fragment:
# %amax : [num_users=1] = call_function[target=torch.ops.aten.amax.default](args = (%arg0_1, [1], True), kwargs = {})
# %sub : [num_users=2] = call_function[target=torch.ops.aten.sub.Tensor](args = (%arg0_1, %amax), kwargs = {})
triton_poi_fused__log_softmax_0 = async_compile.triton('triton_poi_fused__log_softmax_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__log_softmax_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__log_softmax_0(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x0 = xindex % 16
x2 = (xindex // 64)
tmp0 = tl.load(in_ptr0 + (x3), xmask)
tmp1 = tl.load(in_ptr0 + (x0 + (64*x2)), xmask, eviction_policy='evict_last')
tmp2 = tl.load(in_ptr0 + (16 + x0 + (64*x2)), xmask, eviction_policy='evict_last')
tmp4 = tl.load(in_ptr0 + (32 + x0 + (64*x2)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (48 + x0 + (64*x2)), xmask, eviction_policy='evict_last')
tmp3 = triton_helpers.maximum(tmp1, tmp2)
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp7 = triton_helpers.maximum(tmp5, tmp6)
tmp8 = tmp0 - tmp7
tl.store(out_ptr0 + (x3), tmp8, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/ne/cneuoe5ed43ex5ojv524lcm6efihmzp4tx5sn3qedrcityjvt6pz.py
# Topologically Sorted Source Nodes: [log_softmax, inputs], Original ATen: [aten._log_softmax, aten.neg]
# Source node to ATen node mapping:
# inputs => neg
# log_softmax => exp, log, sub_1, sum_1
# Graph fragment:
# %exp : [num_users=1] = call_function[target=torch.ops.aten.exp.default](args = (%sub,), kwargs = {})
# %sum_1 : [num_users=1] = call_function[target=torch.ops.aten.sum.dim_IntList](args = (%exp, [1], True), kwargs = {})
# %log : [num_users=1] = call_function[target=torch.ops.aten.log.default](args = (%sum_1,), kwargs = {})
# %sub_1 : [num_users=1] = call_function[target=torch.ops.aten.sub.Tensor](args = (%sub, %log), kwargs = {})
# %neg : [num_users=4] = call_function[target=torch.ops.aten.neg.default](args = (%sub_1,), kwargs = {})
triton_poi_fused__log_softmax_neg_1 = async_compile.triton('triton_poi_fused__log_softmax_neg_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[256],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused__log_softmax_neg_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 5, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused__log_softmax_neg_1(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x0 = xindex % 16
x2 = (xindex // 64)
tmp0 = tl.load(in_ptr0 + (x3), xmask)
tmp1 = tl.load(in_ptr0 + (x0 + (64*x2)), xmask, eviction_policy='evict_last')
tmp3 = tl.load(in_ptr0 + (16 + x0 + (64*x2)), xmask, eviction_policy='evict_last')
tmp6 = tl.load(in_ptr0 + (32 + x0 + (64*x2)), xmask, eviction_policy='evict_last')
tmp9 = tl.load(in_ptr0 + (48 + x0 + (64*x2)), xmask, eviction_policy='evict_last')
tmp2 = tl_math.exp(tmp1)
tmp4 = tl_math.exp(tmp3)
tmp5 = tmp2 + tmp4
tmp7 = tl_math.exp(tmp6)
tmp8 = tmp5 + tmp7
tmp10 = tl_math.exp(tmp9)
tmp11 = tmp8 + tmp10
tmp12 = tl_math.log(tmp11)
tmp13 = tmp0 - tmp12
tmp14 = -tmp13
tl.store(out_ptr0 + (x3), tmp14, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/hm/chmjikblxn7fogqton2wplz4p5ubizfnslw2y2jskup5crqiglsi.py
# Topologically Sorted Source Nodes: [getitem], Original ATen: [aten.index]
# Source node to ATen node mapping:
# getitem => index
# Graph fragment:
# %index : [num_users=1] = call_function[target=torch.ops.aten.index.Tensor](args = (%neg, [%full_default]), kwargs = {})
triton_poi_fused_index_2 = async_compile.triton('triton_poi_fused_index_2', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[4, 16], tile_hint=TileHint.SQUARE,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_index_2', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_index_2(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 4
xnumel = 16
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x1 = xindex
y0 = yindex
tmp0 = tl.load(in_ptr0 + (x1 + (16*y0)), xmask & ymask, eviction_policy='evict_last')
tl.store(out_ptr0 + (y0 + (4*x1)), tmp0, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/xg/cxgyo7qw6gisgut6u4bocabj7jkiaw3bydso4hzgts7qjaia74xe.py
# Topologically Sorted Source Nodes: [getitem_2], Original ATen: [aten.index]
# Source node to ATen node mapping:
# getitem_2 => index_2
# Graph fragment:
# %index_2 : [num_users=1] = call_function[target=torch.ops.aten.index.Tensor](args = (%neg, [%full_default_2]), kwargs = {})
triton_poi_fused_index_3 = async_compile.triton('triton_poi_fused_index_3', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[4, 16], tile_hint=TileHint.SQUARE,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_index_3', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_index_3(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 4
xnumel = 16
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x1 = xindex
y0 = yindex
tmp0 = tl.load(in_ptr0 + (64 + x1 + (16*y0)), xmask & ymask, eviction_policy='evict_last')
tl.store(out_ptr0 + (y0 + (4*x1)), tmp0, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/iq/ciq2g5midlhlq7225k24quvfyfki7oejp5vufoszhy2i3ktzwl66.py
# Topologically Sorted Source Nodes: [getitem_4], Original ATen: [aten.index]
# Source node to ATen node mapping:
# getitem_4 => index_4
# Graph fragment:
# %index_4 : [num_users=1] = call_function[target=torch.ops.aten.index.Tensor](args = (%neg, [%full_default_4]), kwargs = {})
triton_poi_fused_index_4 = async_compile.triton('triton_poi_fused_index_4', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[4, 16], tile_hint=TileHint.SQUARE,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_index_4', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_index_4(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 4
xnumel = 16
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x1 = xindex
y0 = yindex
tmp0 = tl.load(in_ptr0 + (128 + x1 + (16*y0)), xmask & ymask, eviction_policy='evict_last')
tl.store(out_ptr0 + (y0 + (4*x1)), tmp0, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/uw/cuwcbkvcc6zyei2rc3jzebdq4rl7s3nyyw3nqvslxd7f2o7txngx.py
# Topologically Sorted Source Nodes: [getitem_6], Original ATen: [aten.index]
# Source node to ATen node mapping:
# getitem_6 => index_6
# Graph fragment:
# %index_6 : [num_users=1] = call_function[target=torch.ops.aten.index.Tensor](args = (%neg, [%full_default_6]), kwargs = {})
triton_poi_fused_index_5 = async_compile.triton('triton_poi_fused_index_5', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[4, 16], tile_hint=TileHint.SQUARE,
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32', 3: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 3), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_index_5', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 1, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_index_5(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK : tl.constexpr, XBLOCK : tl.constexpr):
ynumel = 4
xnumel = 16
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x1 = xindex
y0 = yindex
tmp0 = tl.load(in_ptr0 + (192 + x1 + (16*y0)), xmask & ymask, eviction_policy='evict_last')
tl.store(out_ptr0 + (y0 + (4*x1)), tmp0, xmask & ymask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/zz/czzlytqgjwzprkdc572dfyzk5vpiocqqctt74eadeo2aq54ezkih.py
# Topologically Sorted Source Nodes: [truediv, loss, truediv_1, loss_1, truediv_2, loss_2, truediv_3, loss_3], Original ATen: [aten.div, aten.add]
# Source node to ATen node mapping:
# loss => add
# loss_1 => add_1
# loss_2 => add_2
# loss_3 => add_3
# truediv => div
# truediv_1 => div_1
# truediv_2 => div_2
# truediv_3 => div_3
# Graph fragment:
# %div : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%convolution, 16), kwargs = {})
# %add : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%div, 0), kwargs = {})
# %div_1 : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%convolution_1, 16), kwargs = {})
# %add_1 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%add, %div_1), kwargs = {})
# %div_2 : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%convolution_2, 16), kwargs = {})
# %add_2 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%add_1, %div_2), kwargs = {})
# %div_3 : [num_users=1] = call_function[target=torch.ops.aten.div.Tensor](args = (%convolution_3, 16), kwargs = {})
# %add_3 : [num_users=1] = call_function[target=torch.ops.aten.add.Tensor](args = (%add_2, %div_3), kwargs = {})
triton_poi_fused_add_div_6 = async_compile.triton('triton_poi_fused_add_div_6', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[1],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {4: 1}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3), equal_to_1=(4,))]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_div_6', 'mutated_arg_names': ['in_out_ptr0'], 'no_x_dim': False, 'num_load': 4, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_add_div_6(in_out_ptr0, in_ptr0, in_ptr1, in_ptr2, xnumel, XBLOCK : tl.constexpr):
xnumel = 1
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = tl.full([XBLOCK], True, tl.int1)
tmp0 = tl.load(in_ptr0 + (0))
tmp1 = tl.broadcast_to(tmp0, [XBLOCK])
tmp6 = tl.load(in_ptr1 + (0))
tmp7 = tl.broadcast_to(tmp6, [XBLOCK])
tmp10 = tl.load(in_out_ptr0 + (0))
tmp11 = tl.broadcast_to(tmp10, [XBLOCK])
tmp14 = tl.load(in_ptr2 + (0))
tmp15 = tl.broadcast_to(tmp14, [XBLOCK])
tmp2 = 0.0625
tmp3 = tmp1 * tmp2
tmp4 = 0.0
tmp5 = tmp3 + tmp4
tmp8 = tmp7 * tmp2
tmp9 = tmp5 + tmp8
tmp12 = tmp11 * tmp2
tmp13 = tmp9 + tmp12
tmp16 = tmp15 * tmp2
tmp17 = tmp13 + tmp16
tl.store(in_out_ptr0 + (tl.full([XBLOCK], 0, tl.int32)), tmp17, None)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, arg1_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg1_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [log_softmax], Original ATen: [aten._log_softmax]
stream0 = get_raw_stream(0)
triton_poi_fused__log_softmax_0.run(arg0_1, buf0, 256, grid=grid(256), stream=stream0)
del arg0_1
buf1 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
# Topologically Sorted Source Nodes: [log_softmax, inputs], Original ATen: [aten._log_softmax, aten.neg]
triton_poi_fused__log_softmax_neg_1.run(buf0, buf1, 256, grid=grid(256), stream=stream0)
del buf0
buf2 = empty_strided_cuda((1, 4, 4, 4), (64, 1, 16, 4), torch.float32)
# Topologically Sorted Source Nodes: [getitem], Original ATen: [aten.index]
triton_poi_fused_index_2.run(buf1, buf2, 4, 16, grid=grid(4, 16), stream=stream0)
buf3 = empty_strided_cuda((1, 4, 4, 4), (64, 1, 16, 4), torch.float32)
# Topologically Sorted Source Nodes: [getitem_1], Original ATen: [aten.index]
triton_poi_fused_index_2.run(arg1_1, buf3, 4, 16, grid=grid(4, 16), stream=stream0)
# Topologically Sorted Source Nodes: [getitem, getitem_1, conv2d], Original ATen: [aten.index, aten.convolution]
buf4 = extern_kernels.convolution(buf2, buf3, stride=(1, 1), padding=(0, 0), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf4, (1, 1, 1, 1), (1, 1, 1, 1))
buf5 = buf3; del buf3 # reuse
# Topologically Sorted Source Nodes: [getitem_2], Original ATen: [aten.index]
triton_poi_fused_index_3.run(buf1, buf5, 4, 16, grid=grid(4, 16), stream=stream0)
buf6 = buf2; del buf2 # reuse
# Topologically Sorted Source Nodes: [getitem_3], Original ATen: [aten.index]
triton_poi_fused_index_3.run(arg1_1, buf6, 4, 16, grid=grid(4, 16), stream=stream0)
# Topologically Sorted Source Nodes: [getitem_2, getitem_3, conv2d_1], Original ATen: [aten.index, aten.convolution]
buf7 = extern_kernels.convolution(buf5, buf6, stride=(1, 1), padding=(0, 0), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf7, (1, 1, 1, 1), (1, 1, 1, 1))
buf8 = buf6; del buf6 # reuse
# Topologically Sorted Source Nodes: [getitem_4], Original ATen: [aten.index]
triton_poi_fused_index_4.run(buf1, buf8, 4, 16, grid=grid(4, 16), stream=stream0)
buf9 = buf5; del buf5 # reuse
# Topologically Sorted Source Nodes: [getitem_5], Original ATen: [aten.index]
triton_poi_fused_index_4.run(arg1_1, buf9, 4, 16, grid=grid(4, 16), stream=stream0)
# Topologically Sorted Source Nodes: [getitem_4, getitem_5, conv2d_2], Original ATen: [aten.index, aten.convolution]
buf10 = extern_kernels.convolution(buf8, buf9, stride=(1, 1), padding=(0, 0), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf10, (1, 1, 1, 1), (1, 1, 1, 1))
buf11 = buf9; del buf9 # reuse
# Topologically Sorted Source Nodes: [getitem_6], Original ATen: [aten.index]
triton_poi_fused_index_5.run(buf1, buf11, 4, 16, grid=grid(4, 16), stream=stream0)
del buf1
buf12 = buf8; del buf8 # reuse
# Topologically Sorted Source Nodes: [getitem_7], Original ATen: [aten.index]
triton_poi_fused_index_5.run(arg1_1, buf12, 4, 16, grid=grid(4, 16), stream=stream0)
del arg1_1
# Topologically Sorted Source Nodes: [getitem_6, getitem_7, conv2d_3], Original ATen: [aten.index, aten.convolution]
buf13 = extern_kernels.convolution(buf11, buf12, stride=(1, 1), padding=(0, 0), dilation=(1, 1), transposed=False, output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf13, (1, 1, 1, 1), (1, 1, 1, 1))
del buf11
del buf12
buf14 = buf10; del buf10 # reuse
# Topologically Sorted Source Nodes: [truediv, loss, truediv_1, loss_1, truediv_2, loss_2, truediv_3, loss_3], Original ATen: [aten.div, aten.add]
triton_poi_fused_add_div_6.run(buf14, buf4, buf7, buf13, 1, grid=grid(1), stream=stream0)
del buf13
del buf4
del buf7
return (buf14, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
arg1_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([arg0_1, arg1_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.utils
class SoftCrossEntropyLoss2d(nn.Module):
def __init__(self):
super(SoftCrossEntropyLoss2d, self).__init__()
def forward(self, inputs, targets):
loss = 0
inputs = -F.log_softmax(inputs, dim=1)
for index in range(inputs.size()[0]):
loss += F.conv2d(inputs[range(index, index + 1)], targets[range
(index, index + 1)]) / (targets.size()[2] * targets.size()[3])
return loss
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
from torch._inductor.runtime import triton_helpers
from torch._inductor.runtime.triton_helpers import math as tl_math
import torch.nn as nn
import torch.utils
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_poi_fused__log_softmax_0(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.
constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x0 = xindex % 16
x2 = xindex // 64
tmp0 = tl.load(in_ptr0 + x3, xmask)
tmp1 = tl.load(in_ptr0 + (x0 + 64 * x2), xmask, eviction_policy=
'evict_last')
tmp2 = tl.load(in_ptr0 + (16 + x0 + 64 * x2), xmask, eviction_policy=
'evict_last')
tmp4 = tl.load(in_ptr0 + (32 + x0 + 64 * x2), xmask, eviction_policy=
'evict_last')
tmp6 = tl.load(in_ptr0 + (48 + x0 + 64 * x2), xmask, eviction_policy=
'evict_last')
tmp3 = triton_helpers.maximum(tmp1, tmp2)
tmp5 = triton_helpers.maximum(tmp3, tmp4)
tmp7 = triton_helpers.maximum(tmp5, tmp6)
tmp8 = tmp0 - tmp7
tl.store(out_ptr0 + x3, tmp8, xmask)
@triton.jit
def triton_poi_fused__log_softmax_neg_1(in_ptr0, out_ptr0, xnumel, XBLOCK:
tl.constexpr):
xnumel = 256
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x3 = xindex
x0 = xindex % 16
x2 = xindex // 64
tmp0 = tl.load(in_ptr0 + x3, xmask)
tmp1 = tl.load(in_ptr0 + (x0 + 64 * x2), xmask, eviction_policy=
'evict_last')
tmp3 = tl.load(in_ptr0 + (16 + x0 + 64 * x2), xmask, eviction_policy=
'evict_last')
tmp6 = tl.load(in_ptr0 + (32 + x0 + 64 * x2), xmask, eviction_policy=
'evict_last')
tmp9 = tl.load(in_ptr0 + (48 + x0 + 64 * x2), xmask, eviction_policy=
'evict_last')
tmp2 = tl_math.exp(tmp1)
tmp4 = tl_math.exp(tmp3)
tmp5 = tmp2 + tmp4
tmp7 = tl_math.exp(tmp6)
tmp8 = tmp5 + tmp7
tmp10 = tl_math.exp(tmp9)
tmp11 = tmp8 + tmp10
tmp12 = tl_math.log(tmp11)
tmp13 = tmp0 - tmp12
tmp14 = -tmp13
tl.store(out_ptr0 + x3, tmp14, xmask)
@triton.jit
def triton_poi_fused_index_2(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK: tl.
constexpr, XBLOCK: tl.constexpr):
ynumel = 4
xnumel = 16
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x1 = xindex
y0 = yindex
tmp0 = tl.load(in_ptr0 + (x1 + 16 * y0), xmask & ymask, eviction_policy
='evict_last')
tl.store(out_ptr0 + (y0 + 4 * x1), tmp0, xmask & ymask)
@triton.jit
def triton_poi_fused_index_3(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK: tl.
constexpr, XBLOCK: tl.constexpr):
ynumel = 4
xnumel = 16
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x1 = xindex
y0 = yindex
tmp0 = tl.load(in_ptr0 + (64 + x1 + 16 * y0), xmask & ymask,
eviction_policy='evict_last')
tl.store(out_ptr0 + (y0 + 4 * x1), tmp0, xmask & ymask)
@triton.jit
def triton_poi_fused_index_4(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK: tl.
constexpr, XBLOCK: tl.constexpr):
ynumel = 4
xnumel = 16
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x1 = xindex
y0 = yindex
tmp0 = tl.load(in_ptr0 + (128 + x1 + 16 * y0), xmask & ymask,
eviction_policy='evict_last')
tl.store(out_ptr0 + (y0 + 4 * x1), tmp0, xmask & ymask)
@triton.jit
def triton_poi_fused_index_5(in_ptr0, out_ptr0, ynumel, xnumel, YBLOCK: tl.
constexpr, XBLOCK: tl.constexpr):
ynumel = 4
xnumel = 16
yoffset = tl.program_id(1) * YBLOCK
yindex = yoffset + tl.arange(0, YBLOCK)[None, :]
ymask = yindex < ynumel
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:, None]
xmask = xindex < xnumel
x1 = xindex
y0 = yindex
tmp0 = tl.load(in_ptr0 + (192 + x1 + 16 * y0), xmask & ymask,
eviction_policy='evict_last')
tl.store(out_ptr0 + (y0 + 4 * x1), tmp0, xmask & ymask)
@triton.jit
def triton_poi_fused_add_div_6(in_out_ptr0, in_ptr0, in_ptr1, in_ptr2,
xnumel, XBLOCK: tl.constexpr):
xoffset = tl.program_id(0) * XBLOCK
xoffset + tl.arange(0, XBLOCK)[:]
tl.full([XBLOCK], True, tl.int1)
tmp0 = tl.load(in_ptr0 + 0)
tmp1 = tl.broadcast_to(tmp0, [XBLOCK])
tmp6 = tl.load(in_ptr1 + 0)
tmp7 = tl.broadcast_to(tmp6, [XBLOCK])
tmp10 = tl.load(in_out_ptr0 + 0)
tmp11 = tl.broadcast_to(tmp10, [XBLOCK])
tmp14 = tl.load(in_ptr2 + 0)
tmp15 = tl.broadcast_to(tmp14, [XBLOCK])
tmp2 = 0.0625
tmp3 = tmp1 * tmp2
tmp4 = 0.0
tmp5 = tmp3 + tmp4
tmp8 = tmp7 * tmp2
tmp9 = tmp5 + tmp8
tmp12 = tmp11 * tmp2
tmp13 = tmp9 + tmp12
tmp16 = tmp15 * tmp2
tmp17 = tmp13 + tmp16
tl.store(in_out_ptr0 + tl.full([XBLOCK], 0, tl.int32), tmp17, None)
def call(args):
arg0_1, arg1_1 = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
assert_size_stride(arg1_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused__log_softmax_0[grid(256)](arg0_1, buf0, 256,
XBLOCK=128, num_warps=4, num_stages=1)
del arg0_1
buf1 = empty_strided_cuda((4, 4, 4, 4), (64, 16, 4, 1), torch.float32)
triton_poi_fused__log_softmax_neg_1[grid(256)](buf0, buf1, 256,
XBLOCK=256, num_warps=4, num_stages=1)
del buf0
buf2 = empty_strided_cuda((1, 4, 4, 4), (64, 1, 16, 4), torch.float32)
triton_poi_fused_index_2[grid(4, 16)](buf1, buf2, 4, 16, XBLOCK=16,
YBLOCK=4, num_warps=1, num_stages=1)
buf3 = empty_strided_cuda((1, 4, 4, 4), (64, 1, 16, 4), torch.float32)
triton_poi_fused_index_2[grid(4, 16)](arg1_1, buf3, 4, 16, XBLOCK=
16, YBLOCK=4, num_warps=1, num_stages=1)
buf4 = extern_kernels.convolution(buf2, buf3, stride=(1, 1),
padding=(0, 0), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf4, (1, 1, 1, 1), (1, 1, 1, 1))
buf5 = buf3
del buf3
triton_poi_fused_index_3[grid(4, 16)](buf1, buf5, 4, 16, XBLOCK=16,
YBLOCK=4, num_warps=1, num_stages=1)
buf6 = buf2
del buf2
triton_poi_fused_index_3[grid(4, 16)](arg1_1, buf6, 4, 16, XBLOCK=
16, YBLOCK=4, num_warps=1, num_stages=1)
buf7 = extern_kernels.convolution(buf5, buf6, stride=(1, 1),
padding=(0, 0), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf7, (1, 1, 1, 1), (1, 1, 1, 1))
buf8 = buf6
del buf6
triton_poi_fused_index_4[grid(4, 16)](buf1, buf8, 4, 16, XBLOCK=16,
YBLOCK=4, num_warps=1, num_stages=1)
buf9 = buf5
del buf5
triton_poi_fused_index_4[grid(4, 16)](arg1_1, buf9, 4, 16, XBLOCK=
16, YBLOCK=4, num_warps=1, num_stages=1)
buf10 = extern_kernels.convolution(buf8, buf9, stride=(1, 1),
padding=(0, 0), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf10, (1, 1, 1, 1), (1, 1, 1, 1))
buf11 = buf9
del buf9
triton_poi_fused_index_5[grid(4, 16)](buf1, buf11, 4, 16, XBLOCK=16,
YBLOCK=4, num_warps=1, num_stages=1)
del buf1
buf12 = buf8
del buf8
triton_poi_fused_index_5[grid(4, 16)](arg1_1, buf12, 4, 16, XBLOCK=
16, YBLOCK=4, num_warps=1, num_stages=1)
del arg1_1
buf13 = extern_kernels.convolution(buf11, buf12, stride=(1, 1),
padding=(0, 0), dilation=(1, 1), transposed=False,
output_padding=(0, 0), groups=1, bias=None)
assert_size_stride(buf13, (1, 1, 1, 1), (1, 1, 1, 1))
del buf11
del buf12
buf14 = buf10
del buf10
triton_poi_fused_add_div_6[grid(1)](buf14, buf4, buf7, buf13, 1,
XBLOCK=1, num_warps=1, num_stages=1)
del buf13
del buf4
del buf7
return buf14,
class SoftCrossEntropyLoss2dNew(nn.Module):
def __init__(self):
super(SoftCrossEntropyLoss2dNew, self).__init__()
def forward(self, input_0, input_1):
arg0_1 = input_0
arg1_1 = input_1
output = call([arg0_1, arg1_1])
return output[0]
| jameslong95/FasterSeg | SoftCrossEntropyLoss2d | false | 6,928 | [
"MIT"
] | 1 | 872e04964ea46494a6018d9915cee5476e361c27 | https://github.com/jameslong95/FasterSeg/tree/872e04964ea46494a6018d9915cee5476e361c27 | import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.utils
class Model(nn.Module):
def __init__(self):
super().__init__()
def forward(self, inputs, targets):
loss = 0
inputs = -F.log_softmax(inputs, dim=1)
for index in range(inputs.size()[0]):
loss += F.conv2d(inputs[range(index, index + 1)], targets[range
(index, index + 1)]) / (targets.size()[2] * targets.size()[3])
return loss
def get_inputs():
return [torch.rand([4, 4, 4, 4]), torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return []
|
StyleTrack | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/7j/c7jkqpq7g244sdbwoxdaraaqjhcray4b7wkphrkwkxlezb6or6wc.py
# Topologically Sorted Source Nodes: [gram_8], Original ATen: [aten.stack]
# Source node to ATen node mapping:
# gram_8 => cat
# Graph fragment:
# %cat : [num_users=1] = call_function[target=torch.ops.aten.cat.default](args = ([%div, %div_1, %div_2, %div_3],), kwargs = {})
triton_poi_fused_stack_0 = async_compile.triton('triton_poi_fused_stack_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[64],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: '*fp32', 3: '*fp32', 4: '*fp32', 5: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2, 3, 4, 5), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_stack_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 4, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_stack_0(in_ptr0, in_ptr1, in_ptr2, in_ptr3, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = (xindex // 4)
x0 = xindex % 4
x2 = xindex
tmp0 = x1
tmp1 = tl.full([1], 0, tl.int64)
tmp2 = tmp0 >= tmp1
tmp3 = tl.full([1], 4, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = tl.load(in_ptr0 + (x0 + (4*x1)), tmp4 & xmask, other=0.0)
tmp6 = 0.0625
tmp7 = tmp5 * tmp6
tmp8 = tl.full(tmp7.shape, 0.0, tmp7.dtype)
tmp9 = tl.where(tmp4, tmp7, tmp8)
tmp10 = tmp0 >= tmp3
tmp11 = tl.full([1], 8, tl.int64)
tmp12 = tmp0 < tmp11
tmp13 = tmp10 & tmp12
tmp14 = tl.load(in_ptr1 + (x0 + (4*((-4) + x1))), tmp13 & xmask, other=0.0)
tmp15 = tmp14 * tmp6
tmp16 = tl.full(tmp15.shape, 0.0, tmp15.dtype)
tmp17 = tl.where(tmp13, tmp15, tmp16)
tmp18 = tmp0 >= tmp11
tmp19 = tl.full([1], 12, tl.int64)
tmp20 = tmp0 < tmp19
tmp21 = tmp18 & tmp20
tmp22 = tl.load(in_ptr2 + (x0 + (4*((-8) + x1))), tmp21 & xmask, other=0.0)
tmp23 = tmp22 * tmp6
tmp24 = tl.full(tmp23.shape, 0.0, tmp23.dtype)
tmp25 = tl.where(tmp21, tmp23, tmp24)
tmp26 = tmp0 >= tmp19
tmp27 = tl.full([1], 16, tl.int64)
tmp28 = tmp0 < tmp27
tmp29 = tl.load(in_ptr3 + (x0 + (4*((-12) + x1))), tmp26 & xmask, other=0.0)
tmp30 = tmp29 * tmp6
tmp31 = tl.full(tmp30.shape, 0.0, tmp30.dtype)
tmp32 = tl.where(tmp26, tmp30, tmp31)
tmp33 = tl.where(tmp21, tmp25, tmp32)
tmp34 = tl.where(tmp13, tmp17, tmp33)
tmp35 = tl.where(tmp4, tmp9, tmp34)
tl.store(out_ptr0 + (x2), tmp35, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [gram], Original ATen: [aten.mm]
extern_kernels.mm(reinterpret_tensor(arg0_1, (4, 16), (16, 1), 0), reinterpret_tensor(arg0_1, (16, 4), (1, 16), 0), out=buf0)
buf1 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [gram_2], Original ATen: [aten.mm]
extern_kernels.mm(reinterpret_tensor(arg0_1, (4, 16), (16, 1), 64), reinterpret_tensor(arg0_1, (16, 4), (1, 16), 64), out=buf1)
buf2 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [gram_4], Original ATen: [aten.mm]
extern_kernels.mm(reinterpret_tensor(arg0_1, (4, 16), (16, 1), 128), reinterpret_tensor(arg0_1, (16, 4), (1, 16), 128), out=buf2)
buf3 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [gram_6], Original ATen: [aten.mm]
extern_kernels.mm(reinterpret_tensor(arg0_1, (4, 16), (16, 1), 192), reinterpret_tensor(arg0_1, (16, 4), (1, 16), 192), out=buf3)
del arg0_1
buf4 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
# Topologically Sorted Source Nodes: [gram_8], Original ATen: [aten.stack]
stream0 = get_raw_stream(0)
triton_poi_fused_stack_0.run(buf0, buf1, buf2, buf3, buf4, 64, grid=grid(64), stream=stream0)
del buf0
del buf1
del buf2
del buf3
return (reinterpret_tensor(buf4, (4, 4, 4), (16, 4, 1), 0), )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([arg0_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
import torch.nn as nn
def gram_matrix(input):
""" gram matrix for feature assignments """
a, b, c, d = input.size()
allG = []
for i in range(a):
features = input[i].view(b, c * d)
gram = torch.mm(features, features.t())
gram = gram.div(c * d)
allG.append(gram)
return torch.stack(allG)
class StyleTrack(nn.Module):
""" This modules tracks the content image style across many images
for loading (partnered with next module). This is useful for
e.g. maintaining color scheme of the content images
"""
def forward(self, input):
""" forward pass """
gram = gram_matrix(input)
self.value = gram
return input
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {}]
| import torch
from torch._inductor.select_algorithm import extern_kernels
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
import torch.nn as nn
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
@triton.jit
def triton_poi_fused_stack_0(in_ptr0, in_ptr1, in_ptr2, in_ptr3, out_ptr0,
xnumel, XBLOCK: tl.constexpr):
xnumel = 64
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = xindex // 4
x0 = xindex % 4
x2 = xindex
tmp0 = x1
tl.full([1], 0, tl.int64)
tmp3 = tl.full([1], 4, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = tl.load(in_ptr0 + (x0 + 4 * x1), tmp4 & xmask, other=0.0)
tmp6 = 0.0625
tmp7 = tmp5 * tmp6
tmp8 = tl.full(tmp7.shape, 0.0, tmp7.dtype)
tmp9 = tl.where(tmp4, tmp7, tmp8)
tmp10 = tmp0 >= tmp3
tmp11 = tl.full([1], 8, tl.int64)
tmp12 = tmp0 < tmp11
tmp13 = tmp10 & tmp12
tmp14 = tl.load(in_ptr1 + (x0 + 4 * (-4 + x1)), tmp13 & xmask, other=0.0)
tmp15 = tmp14 * tmp6
tmp16 = tl.full(tmp15.shape, 0.0, tmp15.dtype)
tmp17 = tl.where(tmp13, tmp15, tmp16)
tmp18 = tmp0 >= tmp11
tmp19 = tl.full([1], 12, tl.int64)
tmp20 = tmp0 < tmp19
tmp21 = tmp18 & tmp20
tmp22 = tl.load(in_ptr2 + (x0 + 4 * (-8 + x1)), tmp21 & xmask, other=0.0)
tmp23 = tmp22 * tmp6
tmp24 = tl.full(tmp23.shape, 0.0, tmp23.dtype)
tmp25 = tl.where(tmp21, tmp23, tmp24)
tmp26 = tmp0 >= tmp19
tl.full([1], 16, tl.int64)
tmp29 = tl.load(in_ptr3 + (x0 + 4 * (-12 + x1)), tmp26 & xmask, other=0.0)
tmp30 = tmp29 * tmp6
tmp31 = tl.full(tmp30.shape, 0.0, tmp30.dtype)
tmp32 = tl.where(tmp26, tmp30, tmp31)
tmp33 = tl.where(tmp21, tmp25, tmp32)
tmp34 = tl.where(tmp13, tmp17, tmp33)
tmp35 = tl.where(tmp4, tmp9, tmp34)
tl.store(out_ptr0 + x2, tmp35, xmask)
def call(args):
arg0_1, = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(arg0_1, (4, 16), (16, 1), 0),
reinterpret_tensor(arg0_1, (16, 4), (1, 16), 0), out=buf0)
buf1 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(arg0_1, (4, 16), (16, 1), 64),
reinterpret_tensor(arg0_1, (16, 4), (1, 16), 64), out=buf1)
buf2 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(arg0_1, (4, 16), (16, 1), 128),
reinterpret_tensor(arg0_1, (16, 4), (1, 16), 128), out=buf2)
buf3 = empty_strided_cuda((4, 4), (4, 1), torch.float32)
extern_kernels.mm(reinterpret_tensor(arg0_1, (4, 16), (16, 1), 192),
reinterpret_tensor(arg0_1, (16, 4), (1, 16), 192), out=buf3)
del arg0_1
buf4 = empty_strided_cuda((16, 4), (4, 1), torch.float32)
get_raw_stream(0)
triton_poi_fused_stack_0[grid(64)](buf0, buf1, buf2, buf3, buf4, 64,
XBLOCK=64, num_warps=1, num_stages=1)
del buf0
del buf1
del buf2
del buf3
return reinterpret_tensor(buf4, (4, 4, 4), (16, 4, 1), 0),
def gram_matrix(input):
""" gram matrix for feature assignments """
a, b, c, d = input.size()
allG = []
for i in range(a):
features = input[i].view(b, c * d)
gram = torch.mm(features, features.t())
gram = gram.div(c * d)
allG.append(gram)
return torch.stack(allG)
class StyleTrackNew(nn.Module):
""" This modules tracks the content image style across many images
for loading (partnered with next module). This is useful for
e.g. maintaining color scheme of the content images
"""
def forward(self, input_0):
arg0_1 = input_0
output = call([arg0_1])
return output[0]
| jaredaevans/UltrafastNST | StyleTrack | false | 6,929 | [
"MIT"
] | 1 | 6671c6b618ce6bb4920b15f782be962e484a5423 | https://github.com/jaredaevans/UltrafastNST/tree/6671c6b618ce6bb4920b15f782be962e484a5423 | import torch
import torch.nn as nn
def gram_matrix(input):
""" gram matrix for feature assignments """
a, b, c, d = input.size()
allG = []
for i in range(a):
features = input[i].view(b, c * d)
gram = torch.mm(features, features.t())
gram = gram.div(c * d)
allG.append(gram)
return torch.stack(allG)
class Model(nn.Module):
""" This modules tracks the content image style across many images
for loading (partnered with next module). This is useful for
e.g. maintaining color scheme of the content images
"""
def forward(self, input):
""" forward pass """
gram = gram_matrix(input)
self.value = gram
return input
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return []
|
ReflectPad2d | # AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align
from torch import device, empty_strided
from torch._inductor.async_compile import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, grid_combo_kernels, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
empty_strided_xpu = torch._C._dynamo.guards._empty_strided_xpu
reinterpret_tensor = torch._C._dynamo.guards._reinterpret_tensor
alloc_from_pool = torch.ops.inductor._alloc_from_pool
async_compile = AsyncCompile()
# kernel path: runs/run_shard_4/inductor_cache/fo/cfohcgdet6wflivfp6hbudmzhk6yw2j2czym6kwy2yiif6mh3avd.py
# Topologically Sorted Source Nodes: [ins], Original ATen: [aten.cat]
# Source node to ATen node mapping:
# ins => cat
# Graph fragment:
# %cat : [num_users=7] = call_function[target=torch.ops.aten.cat.default](args = ([%slice_12, %slice_20, %slice_28, %arg0_1, %slice_32, %slice_24, %slice_16], 3), kwargs = {})
triton_poi_fused_cat_0 = async_compile.triton('triton_poi_fused_cat_0', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[1024],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_cat_0', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 7, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_cat_0(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 640
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 10
x1 = (xindex // 10)
x2 = xindex
tmp0 = x0
tmp1 = tl.full([1], 0, tl.int64)
tmp2 = tmp0 >= tmp1
tmp3 = tl.full([1], 1, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = tl.load(in_ptr0 + (3 + (4*x1)), tmp4 & xmask, eviction_policy='evict_last', other=0.0)
tmp6 = tmp0 >= tmp3
tmp7 = tl.full([1], 2, tl.int64)
tmp8 = tmp0 < tmp7
tmp9 = tmp6 & tmp8
tmp10 = tl.load(in_ptr0 + (2 + (4*x1)), tmp9 & xmask, eviction_policy='evict_last', other=0.0)
tmp11 = tmp0 >= tmp7
tmp12 = tl.full([1], 3, tl.int64)
tmp13 = tmp0 < tmp12
tmp14 = tmp11 & tmp13
tmp15 = tl.load(in_ptr0 + (1 + (4*x1)), tmp14 & xmask, eviction_policy='evict_last', other=0.0)
tmp16 = tmp0 >= tmp12
tmp17 = tl.full([1], 7, tl.int64)
tmp18 = tmp0 < tmp17
tmp19 = tmp16 & tmp18
tmp20 = tl.load(in_ptr0 + ((4*x1) + ((-3) + x0)), tmp19 & xmask, eviction_policy='evict_last', other=0.0)
tmp21 = tmp0 >= tmp17
tmp22 = tl.full([1], 8, tl.int64)
tmp23 = tmp0 < tmp22
tmp24 = tmp21 & tmp23
tmp25 = tl.load(in_ptr0 + (2 + (4*x1)), tmp24 & xmask, eviction_policy='evict_last', other=0.0)
tmp26 = tmp0 >= tmp22
tmp27 = tl.full([1], 9, tl.int64)
tmp28 = tmp0 < tmp27
tmp29 = tmp26 & tmp28
tmp30 = tl.load(in_ptr0 + (1 + (4*x1)), tmp29 & xmask, eviction_policy='evict_last', other=0.0)
tmp31 = tmp0 >= tmp27
tmp32 = tl.full([1], 10, tl.int64)
tmp33 = tmp0 < tmp32
tmp34 = tl.load(in_ptr0 + (4*x1), tmp31 & xmask, eviction_policy='evict_last', other=0.0)
tmp35 = tl.where(tmp29, tmp30, tmp34)
tmp36 = tl.where(tmp24, tmp25, tmp35)
tmp37 = tl.where(tmp19, tmp20, tmp36)
tmp38 = tl.where(tmp14, tmp15, tmp37)
tmp39 = tl.where(tmp9, tmp10, tmp38)
tmp40 = tl.where(tmp4, tmp5, tmp39)
tl.store(out_ptr0 + (x2), tmp40, xmask)
''', device_str='cuda')
# kernel path: runs/run_shard_4/inductor_cache/lh/clh5qhodzshgdlwsocwfpewyjsmigbgl4r4ofkh3zy7nqeug2s6a.py
# Topologically Sorted Source Nodes: [ins_1], Original ATen: [aten.cat]
# Source node to ATen node mapping:
# ins_1 => cat_1
# Graph fragment:
# %cat_1 : [num_users=1] = call_function[target=torch.ops.aten.cat.default](args = ([%slice_43, %slice_51, %slice_59, %cat, %slice_63, %slice_55, %slice_47], 2), kwargs = {})
triton_poi_fused_cat_1 = async_compile.triton('triton_poi_fused_cat_1', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor
from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor, DeviceProperties
@triton_heuristics.pointwise(
size_hints=[2048],
filename=__file__,
triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': DeviceProperties(type='cuda', index=0, cc=80, major=8, regs_per_multiprocessor=65536, max_threads_per_multi_processor=2048, multi_processor_count=108), 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_cat_1', 'mutated_arg_names': [], 'no_x_dim': False, 'num_load': 7, 'num_reduction': 0, 'backend_hash': 'A9C866B4A14FD3277824029365D703C2427B2E685E54EC9B3EF4ADC8D1EEAC1D', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': None, 'force_disable_caches': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
min_elem_per_thread=0
)
@triton.jit
def triton_poi_fused_cat_1(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
xnumel = 1600
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = (xindex // 10) % 10
x0 = xindex % 10
x2 = (xindex // 100)
x3 = xindex
tmp0 = x1
tmp1 = tl.full([1], 0, tl.int64)
tmp2 = tmp0 >= tmp1
tmp3 = tl.full([1], 1, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = tl.load(in_ptr0 + (30 + x0 + (40*x2)), tmp4 & xmask, eviction_policy='evict_last', other=0.0)
tmp6 = tmp0 >= tmp3
tmp7 = tl.full([1], 2, tl.int64)
tmp8 = tmp0 < tmp7
tmp9 = tmp6 & tmp8
tmp10 = tl.load(in_ptr0 + (20 + x0 + (40*x2)), tmp9 & xmask, eviction_policy='evict_last', other=0.0)
tmp11 = tmp0 >= tmp7
tmp12 = tl.full([1], 3, tl.int64)
tmp13 = tmp0 < tmp12
tmp14 = tmp11 & tmp13
tmp15 = tl.load(in_ptr0 + (10 + x0 + (40*x2)), tmp14 & xmask, eviction_policy='evict_last', other=0.0)
tmp16 = tmp0 >= tmp12
tmp17 = tl.full([1], 7, tl.int64)
tmp18 = tmp0 < tmp17
tmp19 = tmp16 & tmp18
tmp20 = tl.load(in_ptr0 + (x0 + (10*((-3) + x1)) + (40*x2)), tmp19 & xmask, other=0.0)
tmp21 = tmp0 >= tmp17
tmp22 = tl.full([1], 8, tl.int64)
tmp23 = tmp0 < tmp22
tmp24 = tmp21 & tmp23
tmp25 = tl.load(in_ptr0 + (20 + x0 + (40*x2)), tmp24 & xmask, eviction_policy='evict_last', other=0.0)
tmp26 = tmp0 >= tmp22
tmp27 = tl.full([1], 9, tl.int64)
tmp28 = tmp0 < tmp27
tmp29 = tmp26 & tmp28
tmp30 = tl.load(in_ptr0 + (10 + x0 + (40*x2)), tmp29 & xmask, eviction_policy='evict_last', other=0.0)
tmp31 = tmp0 >= tmp27
tmp32 = tl.full([1], 10, tl.int64)
tmp33 = tmp0 < tmp32
tmp34 = tl.load(in_ptr0 + (x0 + (40*x2)), tmp31 & xmask, eviction_policy='evict_last', other=0.0)
tmp35 = tl.where(tmp29, tmp30, tmp34)
tmp36 = tl.where(tmp24, tmp25, tmp35)
tmp37 = tl.where(tmp19, tmp20, tmp36)
tmp38 = tl.where(tmp14, tmp15, tmp37)
tmp39 = tl.where(tmp9, tmp10, tmp38)
tmp40 = tl.where(tmp4, tmp5, tmp39)
tl.store(out_ptr0 + (x3), tmp40, xmask)
''', device_str='cuda')
async_compile.wait(globals())
del async_compile
def call(args):
arg0_1, = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4, 10), (160, 40, 10, 1), torch.float32)
# Topologically Sorted Source Nodes: [ins], Original ATen: [aten.cat]
stream0 = get_raw_stream(0)
triton_poi_fused_cat_0.run(arg0_1, buf0, 640, grid=grid(640), stream=stream0)
del arg0_1
buf1 = empty_strided_cuda((4, 4, 10, 10), (400, 100, 10, 1), torch.float32)
# Topologically Sorted Source Nodes: [ins_1], Original ATen: [aten.cat]
triton_poi_fused_cat_1.run(buf0, buf1, 1600, grid=grid(1600), stream=stream0)
del buf0
return (buf1, )
def benchmark_compiled_module(times=10, repeat=10):
from torch._dynamo.testing import rand_strided
from torch._inductor.utils import print_performance
arg0_1 = rand_strided((4, 4, 4, 4), (64, 16, 4, 1), device='cuda:0', dtype=torch.float32)
fn = lambda: call([arg0_1])
return print_performance(fn, times=times, repeat=repeat)
if __name__ == "__main__":
from torch._inductor.wrapper_benchmark import compiled_module_main
compiled_module_main('None', benchmark_compiled_module)
| import torch
class ReflectPad2d(torch.nn.Module):
""" reflectionpad2d that can be transfered across onnx etc
size : int (the size of padding)
"""
def __init__(self, size):
super().__init__()
self.size = size
def forward(self, ins):
size = self.size
l_list, r_list = [], []
u_list, d_list = [], []
for i in range(size):
left = ins[:, :, :, size - i:size - i + 1]
l_list.append(left)
right = ins[:, :, :, i - size - 1:i - size]
r_list.append(right)
l_list.append(ins)
ins = torch.cat(l_list + r_list[::-1], dim=3)
for i in range(size):
up = ins[:, :, size - i:size - i + 1, :]
u_list.append(up)
down = ins[:, :, i - size - 1:i - size, :]
d_list.append(down)
u_list.append(ins)
ins = torch.cat(u_list + d_list[::-1], dim=2)
return ins
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [[], {'size': 4}]
| import torch
import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid
from torch._C import _cuda_getCurrentRawStream as get_raw_stream
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
@triton.jit
def triton_poi_fused_cat_0(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 640
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x0 = xindex % 10
x1 = xindex // 10
x2 = xindex
tmp0 = x0
tl.full([1], 0, tl.int64)
tmp3 = tl.full([1], 1, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = tl.load(in_ptr0 + (3 + 4 * x1), tmp4 & xmask, eviction_policy=
'evict_last', other=0.0)
tmp6 = tmp0 >= tmp3
tmp7 = tl.full([1], 2, tl.int64)
tmp8 = tmp0 < tmp7
tmp9 = tmp6 & tmp8
tmp10 = tl.load(in_ptr0 + (2 + 4 * x1), tmp9 & xmask, eviction_policy=
'evict_last', other=0.0)
tmp11 = tmp0 >= tmp7
tmp12 = tl.full([1], 3, tl.int64)
tmp13 = tmp0 < tmp12
tmp14 = tmp11 & tmp13
tmp15 = tl.load(in_ptr0 + (1 + 4 * x1), tmp14 & xmask, eviction_policy=
'evict_last', other=0.0)
tmp16 = tmp0 >= tmp12
tmp17 = tl.full([1], 7, tl.int64)
tmp18 = tmp0 < tmp17
tmp19 = tmp16 & tmp18
tmp20 = tl.load(in_ptr0 + (4 * x1 + (-3 + x0)), tmp19 & xmask,
eviction_policy='evict_last', other=0.0)
tmp21 = tmp0 >= tmp17
tmp22 = tl.full([1], 8, tl.int64)
tmp23 = tmp0 < tmp22
tmp24 = tmp21 & tmp23
tmp25 = tl.load(in_ptr0 + (2 + 4 * x1), tmp24 & xmask, eviction_policy=
'evict_last', other=0.0)
tmp26 = tmp0 >= tmp22
tmp27 = tl.full([1], 9, tl.int64)
tmp28 = tmp0 < tmp27
tmp29 = tmp26 & tmp28
tmp30 = tl.load(in_ptr0 + (1 + 4 * x1), tmp29 & xmask, eviction_policy=
'evict_last', other=0.0)
tmp31 = tmp0 >= tmp27
tl.full([1], 10, tl.int64)
tmp34 = tl.load(in_ptr0 + 4 * x1, tmp31 & xmask, eviction_policy=
'evict_last', other=0.0)
tmp35 = tl.where(tmp29, tmp30, tmp34)
tmp36 = tl.where(tmp24, tmp25, tmp35)
tmp37 = tl.where(tmp19, tmp20, tmp36)
tmp38 = tl.where(tmp14, tmp15, tmp37)
tmp39 = tl.where(tmp9, tmp10, tmp38)
tmp40 = tl.where(tmp4, tmp5, tmp39)
tl.store(out_ptr0 + x2, tmp40, xmask)
@triton.jit
def triton_poi_fused_cat_1(in_ptr0, out_ptr0, xnumel, XBLOCK: tl.constexpr):
xnumel = 1600
xoffset = tl.program_id(0) * XBLOCK
xindex = xoffset + tl.arange(0, XBLOCK)[:]
xmask = xindex < xnumel
x1 = xindex // 10 % 10
x0 = xindex % 10
x2 = xindex // 100
x3 = xindex
tmp0 = x1
tl.full([1], 0, tl.int64)
tmp3 = tl.full([1], 1, tl.int64)
tmp4 = tmp0 < tmp3
tmp5 = tl.load(in_ptr0 + (30 + x0 + 40 * x2), tmp4 & xmask,
eviction_policy='evict_last', other=0.0)
tmp6 = tmp0 >= tmp3
tmp7 = tl.full([1], 2, tl.int64)
tmp8 = tmp0 < tmp7
tmp9 = tmp6 & tmp8
tmp10 = tl.load(in_ptr0 + (20 + x0 + 40 * x2), tmp9 & xmask,
eviction_policy='evict_last', other=0.0)
tmp11 = tmp0 >= tmp7
tmp12 = tl.full([1], 3, tl.int64)
tmp13 = tmp0 < tmp12
tmp14 = tmp11 & tmp13
tmp15 = tl.load(in_ptr0 + (10 + x0 + 40 * x2), tmp14 & xmask,
eviction_policy='evict_last', other=0.0)
tmp16 = tmp0 >= tmp12
tmp17 = tl.full([1], 7, tl.int64)
tmp18 = tmp0 < tmp17
tmp19 = tmp16 & tmp18
tmp20 = tl.load(in_ptr0 + (x0 + 10 * (-3 + x1) + 40 * x2), tmp19 &
xmask, other=0.0)
tmp21 = tmp0 >= tmp17
tmp22 = tl.full([1], 8, tl.int64)
tmp23 = tmp0 < tmp22
tmp24 = tmp21 & tmp23
tmp25 = tl.load(in_ptr0 + (20 + x0 + 40 * x2), tmp24 & xmask,
eviction_policy='evict_last', other=0.0)
tmp26 = tmp0 >= tmp22
tmp27 = tl.full([1], 9, tl.int64)
tmp28 = tmp0 < tmp27
tmp29 = tmp26 & tmp28
tmp30 = tl.load(in_ptr0 + (10 + x0 + 40 * x2), tmp29 & xmask,
eviction_policy='evict_last', other=0.0)
tmp31 = tmp0 >= tmp27
tl.full([1], 10, tl.int64)
tmp34 = tl.load(in_ptr0 + (x0 + 40 * x2), tmp31 & xmask,
eviction_policy='evict_last', other=0.0)
tmp35 = tl.where(tmp29, tmp30, tmp34)
tmp36 = tl.where(tmp24, tmp25, tmp35)
tmp37 = tl.where(tmp19, tmp20, tmp36)
tmp38 = tl.where(tmp14, tmp15, tmp37)
tmp39 = tl.where(tmp9, tmp10, tmp38)
tmp40 = tl.where(tmp4, tmp5, tmp39)
tl.store(out_ptr0 + x3, tmp40, xmask)
def call(args):
arg0_1, = args
args.clear()
assert_size_stride(arg0_1, (4, 4, 4, 4), (64, 16, 4, 1))
with torch.cuda._DeviceGuard(0):
torch.cuda.set_device(0)
buf0 = empty_strided_cuda((4, 4, 4, 10), (160, 40, 10, 1), torch.
float32)
get_raw_stream(0)
triton_poi_fused_cat_0[grid(640)](arg0_1, buf0, 640, XBLOCK=256,
num_warps=4, num_stages=1)
del arg0_1
buf1 = empty_strided_cuda((4, 4, 10, 10), (400, 100, 10, 1), torch.
float32)
triton_poi_fused_cat_1[grid(1600)](buf0, buf1, 1600, XBLOCK=128,
num_warps=4, num_stages=1)
del buf0
return buf1,
class ReflectPad2dNew(torch.nn.Module):
""" reflectionpad2d that can be transfered across onnx etc
size : int (the size of padding)
"""
def __init__(self, size):
super().__init__()
self.size = size
def forward(self, input_0):
arg0_1 = input_0
output = call([arg0_1])
return output[0]
| jaredaevans/UltrafastNST | ReflectPad2d | false | 6,930 | [
"MIT"
] | 1 | 6671c6b618ce6bb4920b15f782be962e484a5423 | https://github.com/jaredaevans/UltrafastNST/tree/6671c6b618ce6bb4920b15f782be962e484a5423 | import torch
class Model(torch.nn.Module):
""" reflectionpad2d that can be transfered across onnx etc
size : int (the size of padding)
"""
def __init__(self, size):
super().__init__()
self.size = size
def forward(self, ins):
size = self.size
l_list, r_list = [], []
u_list, d_list = [], []
for i in range(size):
left = ins[:, :, :, size - i:size - i + 1]
l_list.append(left)
right = ins[:, :, :, i - size - 1:i - size]
r_list.append(right)
l_list.append(ins)
ins = torch.cat(l_list + r_list[::-1], dim=3)
for i in range(size):
up = ins[:, :, size - i:size - i + 1, :]
u_list.append(up)
down = ins[:, :, i - size - 1:i - size, :]
d_list.append(down)
u_list.append(ins)
ins = torch.cat(u_list + d_list[::-1], dim=2)
return ins
def get_inputs():
return [torch.rand([4, 4, 4, 4])]
def get_init_inputs():
return [4]
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.