Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
@@ -1,4 +1,30 @@
|
|
1 |
import gradio as gr
|
|
|
|
|
|
|
|
|
2 |
|
3 |
-
|
4 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
import gradio as gr
|
2 |
+
from huggingface_hub import snapshot_download
|
3 |
+
from accelerate.utils import BnbQuantizationConfig
|
4 |
+
from accelerate.utils import load_and_quantize_model
|
5 |
+
from accelerate import Accelerator
|
6 |
|
7 |
+
model_path="marcsun13/gpt2-xl-linear-sharded"
|
8 |
+
|
9 |
+
def quantize(model_path=model_path):
|
10 |
+
print("1")
|
11 |
+
weights_location = snapshot_download(repo_id=f"{model_path}")
|
12 |
+
print("2")
|
13 |
+
bnb_quantization_config = BnbQuantizationConfig(load_in_8bit=True, llm_int8_threshold = 6)
|
14 |
+
#bnb_quantization_config = BnbQuantizationConfig(load_in_4bit=True, bnb_4bit_compute_dtype=torch.bfloat16, bnb_4bit_use_double_quant=True, bnb_4bit_quant_type="nf4")
|
15 |
+
print("3")
|
16 |
+
quantized_model = load_and_quantize_model(empty_model, weights_location=weights_location, bnb_quantization_config=bnb_quantization_config, device_map = "auto")
|
17 |
+
print("4")
|
18 |
+
accelerate = Accelerator()
|
19 |
+
print("5")
|
20 |
+
new_weights_location = "./model"
|
21 |
+
print("6")
|
22 |
+
accelerate.save_model(quantized_model, new_weights_location)
|
23 |
+
print("7")
|
24 |
+
quantized_model_from_saved = load_and_quantize_model(empty_model, weights_location=new_weights_location, bnb_quantization_config=bnb_quantization_config, device_map = "auto")
|
25 |
+
print("Done")
|
26 |
+
|
27 |
+
with gr.Blocks() as app:
|
28 |
+
btn=gr.Button()
|
29 |
+
btn.click(quantize,None,None)
|
30 |
+
app.launch()
|