Hugging Face's logo

qualcomm
/
ResNet101

Image Classification
PyTorch
LiteRT
ONNX
backbone
android
Model card Files Community
qaihm-bot commited on
Commit
56c1676
·
verified ·
1 Parent(s): ade7da7

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +43 -25
README.md CHANGED
@@ -35,29 +35,31 @@ 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
- | ResNet101 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 3.394 ms | 0 - 2 MB | FP16 | NPU | [ResNet101.tflite](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.tflite) |
39
- | ResNet101 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 3.496 ms | 0 - 79 MB | FP16 | NPU | [ResNet101.so](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.so) |
40
- | ResNet101 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 3.506 ms | 0 - 98 MB | FP16 | NPU | [ResNet101.onnx](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.onnx) |
41
- | ResNet101 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 2.424 ms | 0 - 114 MB | FP16 | NPU | [ResNet101.tflite](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.tflite) |
42
- | ResNet101 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 2.509 ms | 1 - 35 MB | FP16 | NPU | [ResNet101.so](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.so) |
43
- | ResNet101 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 2.557 ms | 0 - 119 MB | FP16 | NPU | [ResNet101.onnx](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.onnx) |
44
- | ResNet101 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 2.354 ms | 0 - 43 MB | FP16 | NPU | [ResNet101.tflite](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.tflite) |
45
- | ResNet101 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 2.404 ms | 0 - 32 MB | FP16 | NPU | Use Export Script |
46
- | ResNet101 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 2.532 ms | 0 - 45 MB | FP16 | NPU | [ResNet101.onnx](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.onnx) |
47
- | ResNet101 | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 3.382 ms | 0 - 16 MB | FP16 | NPU | [ResNet101.tflite](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.tflite) |
48
- | ResNet101 | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 3.279 ms | 1 - 2 MB | FP16 | NPU | Use Export Script |
49
- | ResNet101 | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 3.405 ms | 0 - 2 MB | FP16 | NPU | [ResNet101.tflite](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.tflite) |
50
- | ResNet101 | SA8255 (Proxy) | SA8255P Proxy | QNN | 3.275 ms | 0 - 1 MB | FP16 | NPU | Use Export Script |
51
- | ResNet101 | SA8775 (Proxy) | SA8775P Proxy | TFLITE | 3.404 ms | 0 - 2 MB | FP16 | NPU | [ResNet101.tflite](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.tflite) |
52
- | ResNet101 | SA8775 (Proxy) | SA8775P Proxy | QNN | 3.314 ms | 1 - 2 MB | FP16 | NPU | Use Export Script |
53
- | ResNet101 | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 3.397 ms | 0 - 2 MB | FP16 | NPU | [ResNet101.tflite](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.tflite) |
54
- | ResNet101 | SA8650 (Proxy) | SA8650P Proxy | QNN | 3.317 ms | 1 - 2 MB | FP16 | NPU | Use Export Script |
55
- | ResNet101 | SA8295P ADP | SA8295P | TFLITE | 5.71 ms | 0 - 23 MB | FP16 | NPU | [ResNet101.tflite](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.tflite) |
56
- | ResNet101 | SA8295P ADP | SA8295P | QNN | 5.719 ms | 1 - 6 MB | FP16 | NPU | Use Export Script |
57
- | ResNet101 | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 4.772 ms | 0 - 93 MB | FP16 | NPU | [ResNet101.tflite](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.tflite) |
58
- | ResNet101 | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 4.883 ms | 1 - 25 MB | FP16 | NPU | Use Export Script |
59
- | ResNet101 | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 3.469 ms | 1 - 1 MB | FP16 | NPU | Use Export Script |
60
- | ResNet101 | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 3.554 ms | 86 - 86 MB | FP16 | NPU | [ResNet101.onnx](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.onnx) |
 
 
61
 
62
 
63
 
@@ -122,7 +124,7 @@ ResNet101
122
  Device : Samsung Galaxy S23 (13)
123
  Runtime : TFLITE
124
  Estimated inference time (ms) : 3.4
125
- Estimated peak memory usage (MB): [0, 2]
126
  Total # Ops : 147
127
  Compute Unit(s) : NPU (147 ops)
128
  ```
@@ -143,13 +145,29 @@ in memory using the `jit.trace` and then call the `submit_compile_job` API.
143
  import torch
144
 
145
  import qai_hub as hub
146
- from qai_hub_models.models.resnet101 import
147
 
148
  # Load the model
 
149
 
150
  # Device
151
  device = hub.Device("Samsung Galaxy S23")
152
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
153
 
154
  ```
155
 
 
35
 
36
  | Model | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
37
  |---|---|---|---|---|---|---|---|---|
38
+ | ResNet101 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 3.424 ms | 0 - 280 MB | FP16 | NPU | [ResNet101.tflite](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.tflite) |
39
+ | ResNet101 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 3.455 ms | 1 - 200 MB | FP16 | NPU | [ResNet101.so](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.so) |
40
+ | ResNet101 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 3.493 ms | 0 - 98 MB | FP16 | NPU | [ResNet101.onnx](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.onnx) |
41
+ | ResNet101 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 2.439 ms | 0 - 34 MB | FP16 | NPU | [ResNet101.tflite](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.tflite) |
42
+ | ResNet101 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 2.495 ms | 1 - 35 MB | FP16 | NPU | [ResNet101.so](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.so) |
43
+ | ResNet101 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 2.593 ms | 0 - 117 MB | FP16 | NPU | [ResNet101.onnx](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.onnx) |
44
+ | ResNet101 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 2.346 ms | 0 - 33 MB | FP16 | NPU | [ResNet101.tflite](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.tflite) |
45
+ | ResNet101 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 2.403 ms | 0 - 31 MB | FP16 | NPU | Use Export Script |
46
+ | ResNet101 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 2.518 ms | 0 - 46 MB | FP16 | NPU | [ResNet101.onnx](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.onnx) |
47
+ | ResNet101 | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 3.379 ms | 0 - 290 MB | FP16 | NPU | [ResNet101.tflite](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.tflite) |
48
+ | ResNet101 | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 3.29 ms | 1 - 2 MB | FP16 | NPU | Use Export Script |
49
+ | ResNet101 | SA7255P ADP | SA7255P | TFLITE | 135.286 ms | 0 - 31 MB | FP16 | NPU | [ResNet101.tflite](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.tflite) |
50
+ | ResNet101 | SA7255P ADP | SA7255P | QNN | 136.421 ms | 1 - 6 MB | FP16 | NPU | Use Export Script |
51
+ | ResNet101 | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 3.368 ms | 0 - 269 MB | FP16 | NPU | [ResNet101.tflite](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.tflite) |
52
+ | ResNet101 | SA8255 (Proxy) | SA8255P Proxy | QNN | 3.334 ms | 1 - 2 MB | FP16 | NPU | Use Export Script |
53
+ | ResNet101 | SA8295P ADP | SA8295P | TFLITE | 5.7 ms | 0 - 19 MB | FP16 | NPU | [ResNet101.tflite](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.tflite) |
54
+ | ResNet101 | SA8295P ADP | SA8295P | QNN | 6.342 ms | 1 - 6 MB | FP16 | NPU | Use Export Script |
55
+ | ResNet101 | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 3.381 ms | 0 - 269 MB | FP16 | NPU | [ResNet101.tflite](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.tflite) |
56
+ | ResNet101 | SA8650 (Proxy) | SA8650P Proxy | QNN | 3.282 ms | 0 - 1 MB | FP16 | NPU | Use Export Script |
57
+ | ResNet101 | SA8775P ADP | SA8775P | TFLITE | 6.643 ms | 0 - 31 MB | FP16 | NPU | [ResNet101.tflite](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.tflite) |
58
+ | ResNet101 | SA8775P ADP | SA8775P | QNN | 6.74 ms | 0 - 6 MB | FP16 | NPU | Use Export Script |
59
+ | ResNet101 | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 4.762 ms | 0 - 24 MB | FP16 | NPU | [ResNet101.tflite](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.tflite) |
60
+ | ResNet101 | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 4.855 ms | 1 - 22 MB | FP16 | NPU | Use Export Script |
61
+ | ResNet101 | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 3.461 ms | 1 - 1 MB | FP16 | NPU | Use Export Script |
62
+ | ResNet101 | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 3.48 ms | 87 - 87 MB | FP16 | NPU | [ResNet101.onnx](https://huggingface.co/qualcomm/ResNet101/blob/main/ResNet101.onnx) |
63
 
64
 
65
 
 
124
  Device : Samsung Galaxy S23 (13)
125
  Runtime : TFLITE
126
  Estimated inference time (ms) : 3.4
127
+ Estimated peak memory usage (MB): [0, 280]
128
  Total # Ops : 147
129
  Compute Unit(s) : NPU (147 ops)
130
  ```
 
145
  import torch
146
 
147
  import qai_hub as hub
148
+ from qai_hub_models.models.resnet101 import Model
149
 
150
  # Load the model
151
+ torch_model = Model.from_pretrained()
152
 
153
  # Device
154
  device = hub.Device("Samsung Galaxy S23")
155
 
156
+ # Trace model
157
+ input_shape = torch_model.get_input_spec()
158
+ sample_inputs = torch_model.sample_inputs()
159
+
160
+ pt_model = torch.jit.trace(torch_model, [torch.tensor(data[0]) for _, data in sample_inputs.items()])
161
+
162
+ # Compile model on a specific device
163
+ compile_job = hub.submit_compile_job(
164
+ model=pt_model,
165
+ device=device,
166
+ input_specs=torch_model.get_input_spec(),
167
+ )
168
+
169
+ # Get target model to run on-device
170
+ target_model = compile_job.get_target_model()
171
 
172
  ```
173