Hugging Face's logo

qualcomm
/
QuickSRNetLarge-Quantized

Image-to-Image
PyTorch
LiteRT
ONNX
quantized
android
Model card Files Community
bhushans commited on
Commit
23f44e4
·
verified ·
1 Parent(s): e62703b

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +45 -28
README.md CHANGED
@@ -35,32 +35,33 @@ More details on model performance across various devices, can be found
35
 
36
  | Model | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
37
  |---|---|---|---|---|---|---|---|---|
38
- | QuickSRNetLarge-Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 1.432 ms | 0 - 1 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.tflite) |
39
- | QuickSRNetLarge-Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 0.908 ms | 0 - 8 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.so](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.so) |
40
- | QuickSRNetLarge-Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 19.669 ms | 23 - 25 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.onnx](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.onnx) |
41
- | QuickSRNetLarge-Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 1.12 ms | 0 - 28 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.tflite) |
42
- | QuickSRNetLarge-Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 0.64 ms | 0 - 14 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.so](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.so) |
43
- | QuickSRNetLarge-Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 13.988 ms | 25 - 104 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.onnx](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.onnx) |
44
- | QuickSRNetLarge-Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 1.334 ms | 0 - 20 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.tflite) |
45
- | QuickSRNetLarge-Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 0.652 ms | 0 - 11 MB | INT8 | NPU | Use Export Script |
46
- | QuickSRNetLarge-Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 13.222 ms | 27 - 75 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.onnx](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.onnx) |
47
- | QuickSRNetLarge-Quantized | RB3 Gen 2 (Proxy) | QCS6490 Proxy | TFLITE | 4.11 ms | 2 - 22 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.tflite) |
48
- | QuickSRNetLarge-Quantized | RB3 Gen 2 (Proxy) | QCS6490 Proxy | QNN | 3.149 ms | 0 - 8 MB | INT8 | NPU | Use Export Script |
49
- | QuickSRNetLarge-Quantized | RB5 (Proxy) | QCS8250 Proxy | TFLITE | 38.391 ms | 2 - 4 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.tflite) |
50
- | QuickSRNetLarge-Quantized | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 1.432 ms | 0 - 1 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.tflite) |
51
- | QuickSRNetLarge-Quantized | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 0.678 ms | 0 - 1 MB | INT8 | NPU | Use Export Script |
52
- | QuickSRNetLarge-Quantized | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 1.453 ms | 2 - 3 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.tflite) |
53
- | QuickSRNetLarge-Quantized | SA8255 (Proxy) | SA8255P Proxy | QNN | 0.673 ms | 0 - 1 MB | INT8 | NPU | Use Export Script |
54
- | QuickSRNetLarge-Quantized | SA8775 (Proxy) | SA8775P Proxy | TFLITE | 1.447 ms | 1 - 2 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.tflite) |
55
- | QuickSRNetLarge-Quantized | SA8775 (Proxy) | SA8775P Proxy | QNN | 0.679 ms | 0 - 1 MB | INT8 | NPU | Use Export Script |
56
- | QuickSRNetLarge-Quantized | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 1.45 ms | 0 - 3 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.tflite) |
57
- | QuickSRNetLarge-Quantized | SA8650 (Proxy) | SA8650P Proxy | QNN | 0.681 ms | 0 - 2 MB | INT8 | NPU | Use Export Script |
58
- | QuickSRNetLarge-Quantized | SA8295P ADP | SA8295P | TFLITE | 2.58 ms | 0 - 19 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.tflite) |
59
- | QuickSRNetLarge-Quantized | SA8295P ADP | SA8295P | QNN | 1.471 ms | 0 - 6 MB | INT8 | NPU | Use Export Script |
60
- | QuickSRNetLarge-Quantized | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 1.885 ms | 0 - 29 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.tflite) |
61
- | QuickSRNetLarge-Quantized | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 1.048 ms | 0 - 13 MB | INT8 | NPU | Use Export Script |
62
- | QuickSRNetLarge-Quantized | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 0.794 ms | 0 - 0 MB | INT8 | NPU | Use Export Script |
63
- | QuickSRNetLarge-Quantized | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 17.516 ms | 33 - 33 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.onnx](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.onnx) |
 
64
 
65
 
66
 
@@ -126,7 +127,7 @@ QuickSRNetLarge-Quantized
126
  Device : Samsung Galaxy S23 (13)
127
  Runtime : TFLITE
128
  Estimated inference time (ms) : 1.4
129
- Estimated peak memory usage (MB): [0, 1]
130
  Total # Ops : 33
131
  Compute Unit(s) : NPU (30 ops) CPU (3 ops)
132
  ```
@@ -147,13 +148,29 @@ in memory using the `jit.trace` and then call the `submit_compile_job` API.
147
  import torch
148
 
149
  import qai_hub as hub
150
- from qai_hub_models.models.quicksrnetlarge_quantized import
151
 
152
  # Load the model
 
153
 
154
  # Device
155
  device = hub.Device("Samsung Galaxy S23")
156
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
157
 
158
  ```
159
 
 
35
 
36
  | Model | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
37
  |---|---|---|---|---|---|---|---|---|
38
+ | QuickSRNetLarge-Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 1.439 ms | 0 - 5 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.tflite) |
39
+ | QuickSRNetLarge-Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 0.91 ms | 0 - 8 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.so](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.so) |
40
+ | QuickSRNetLarge-Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 16.891 ms | 24 - 235 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.onnx](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.onnx) |
41
+ | QuickSRNetLarge-Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 1.105 ms | 0 - 14 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.tflite) |
42
+ | QuickSRNetLarge-Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 0.64 ms | 0 - 12 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.so](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.so) |
43
+ | QuickSRNetLarge-Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 14.179 ms | 25 - 102 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.onnx](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.onnx) |
44
+ | QuickSRNetLarge-Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 1.296 ms | 0 - 11 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.tflite) |
45
+ | QuickSRNetLarge-Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 0.553 ms | 0 - 11 MB | INT8 | NPU | Use Export Script |
46
+ | QuickSRNetLarge-Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 13.266 ms | 22 - 71 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.onnx](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.onnx) |
47
+ | QuickSRNetLarge-Quantized | RB3 Gen 2 (Proxy) | QCS6490 Proxy | TFLITE | 4.171 ms | 2 - 15 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.tflite) |
48
+ | QuickSRNetLarge-Quantized | RB3 Gen 2 (Proxy) | QCS6490 Proxy | QNN | 3.078 ms | 0 - 7 MB | INT8 | NPU | Use Export Script |
49
+ | QuickSRNetLarge-Quantized | RB5 (Proxy) | QCS8250 Proxy | TFLITE | 37.582 ms | 2 - 9 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.tflite) |
50
+ | QuickSRNetLarge-Quantized | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 1.436 ms | 0 - 4 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.tflite) |
51
+ | QuickSRNetLarge-Quantized | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 0.679 ms | 0 - 1 MB | INT8 | NPU | Use Export Script |
52
+ | QuickSRNetLarge-Quantized | SA7255P ADP | SA7255P | TFLITE | 14.18 ms | 2 - 12 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.tflite) |
53
+ | QuickSRNetLarge-Quantized | SA7255P ADP | SA7255P | QNN | 11.883 ms | 0 - 6 MB | INT8 | NPU | Use Export Script |
54
+ | QuickSRNetLarge-Quantized | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 1.446 ms | 0 - 79 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.tflite) |
55
+ | QuickSRNetLarge-Quantized | SA8255 (Proxy) | SA8255P Proxy | QNN | 0.682 ms | 0 - 1 MB | INT8 | NPU | Use Export Script |
56
+ | QuickSRNetLarge-Quantized | SA8295P ADP | SA8295P | TFLITE | 2.491 ms | 0 - 10 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.tflite) |
57
+ | QuickSRNetLarge-Quantized | SA8295P ADP | SA8295P | QNN | 1.48 ms | 0 - 6 MB | INT8 | NPU | Use Export Script |
58
+ | QuickSRNetLarge-Quantized | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 1.451 ms | 0 - 4 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.tflite) |
59
+ | QuickSRNetLarge-Quantized | SA8650 (Proxy) | SA8650P Proxy | QNN | 0.681 ms | 0 - 1 MB | INT8 | NPU | Use Export Script |
60
+ | QuickSRNetLarge-Quantized | SA8775P ADP | SA8775P | TFLITE | 2.577 ms | 1 - 10 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.tflite) |
61
+ | QuickSRNetLarge-Quantized | SA8775P ADP | SA8775P | QNN | 1.249 ms | 0 - 6 MB | INT8 | NPU | Use Export Script |
62
+ | QuickSRNetLarge-Quantized | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 2.317 ms | 0 - 14 MB | INT8 | NPU | [QuickSRNetLarge-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetLarge-Quantized/blob/main/QuickSRNetLarge-Quantized.tflite) |
63
+ | QuickSRNetLarge-Quantized | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 1.06 ms | 0 - 14 MB | INT8 | NPU | Use Export Script |
64
+ | QuickSRNetLarge-Quantized | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 0.798 ms | 0 - 0 MB | INT8 | NPU | Use Export Script |
65
 
66
 
67
 
 
127
  Device : Samsung Galaxy S23 (13)
128
  Runtime : TFLITE
129
  Estimated inference time (ms) : 1.4
130
+ Estimated peak memory usage (MB): [0, 5]
131
  Total # Ops : 33
132
  Compute Unit(s) : NPU (30 ops) CPU (3 ops)
133
  ```
 
148
  import torch
149
 
150
  import qai_hub as hub
151
+ from qai_hub_models.models.quicksrnetlarge_quantized import Model
152
 
153
  # Load the model
154
+ torch_model = Model.from_pretrained()
155
 
156
  # Device
157
  device = hub.Device("Samsung Galaxy S23")
158
 
159
+ # Trace model
160
+ input_shape = torch_model.get_input_spec()
161
+ sample_inputs = torch_model.sample_inputs()
162
+
163
+ pt_model = torch.jit.trace(torch_model, [torch.tensor(data[0]) for _, data in sample_inputs.items()])
164
+
165
+ # Compile model on a specific device
166
+ compile_job = hub.submit_compile_job(
167
+ model=pt_model,
168
+ device=device,
169
+ input_specs=torch_model.get_input_spec(),
170
+ )
171
+
172
+ # Get target model to run on-device
173
+ target_model = compile_job.get_target_model()
174
 
175
  ```
176