Hugging Face's logo

qualcomm
/
ResNeXt101Quantized

Image Classification
PyTorch
TF Lite
ONNX
backbone
quantized
android
Model card Files Community
qaihm-bot commited on
Commit
6e9f160
·
verified ·
1 Parent(s): c4cfcae

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +33 -31
README.md CHANGED
@@ -36,41 +36,42 @@ More details on model performance across various devices, can be found
36
 
37
  | Model | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
38
  |---|---|---|---|---|---|---|---|---|
39
- | ResNeXt101Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 2.75 ms | 0 - 33 MB | INT8 | NPU | [ResNeXt101Quantized.tflite](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.tflite) |
40
- | ResNeXt101Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 3.024 ms | 0 - 28 MB | INT8 | NPU | [ResNeXt101Quantized.so](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.so) |
41
- | ResNeXt101Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 3.829 ms | 0 - 175 MB | INT8 | NPU | [ResNeXt101Quantized.onnx](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.onnx) |
42
- | ResNeXt101Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 2.033 ms | 0 - 97 MB | INT8 | NPU | [ResNeXt101Quantized.tflite](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.tflite) |
43
- | ResNeXt101Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 2.292 ms | 0 - 94 MB | INT8 | NPU | [ResNeXt101Quantized.so](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.so) |
44
- | ResNeXt101Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 2.888 ms | 0 - 117 MB | INT8 | NPU | [ResNeXt101Quantized.onnx](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.onnx) |
45
- | ResNeXt101Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 1.782 ms | 0 - 89 MB | INT8 | NPU | [ResNeXt101Quantized.tflite](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.tflite) |
46
- | ResNeXt101Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 2.275 ms | 0 - 93 MB | INT8 | NPU | Use Export Script |
47
- | ResNeXt101Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 2.982 ms | 0 - 111 MB | INT8 | NPU | [ResNeXt101Quantized.onnx](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.onnx) |
48
- | ResNeXt101Quantized | RB3 Gen 2 (Proxy) | QCS6490 Proxy | TFLITE | 10.224 ms | 0 - 99 MB | INT8 | NPU | [ResNeXt101Quantized.tflite](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.tflite) |
49
- | ResNeXt101Quantized | RB3 Gen 2 (Proxy) | QCS6490 Proxy | QNN | 14.373 ms | 0 - 12 MB | INT8 | NPU | Use Export Script |
50
- | ResNeXt101Quantized | RB5 (Proxy) | QCS8250 Proxy | TFLITE | 129.145 ms | 0 - 293 MB | INT8 | GPU | [ResNeXt101Quantized.tflite](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.tflite) |
51
- | ResNeXt101Quantized | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 2.701 ms | 0 - 34 MB | INT8 | NPU | [ResNeXt101Quantized.tflite](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.tflite) |
52
- | ResNeXt101Quantized | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 2.949 ms | 0 - 3 MB | INT8 | NPU | Use Export Script |
53
- | ResNeXt101Quantized | SA7255P ADP | SA7255P | TFLITE | 34.317 ms | 0 - 87 MB | INT8 | NPU | [ResNeXt101Quantized.tflite](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.tflite) |
54
- | ResNeXt101Quantized | SA7255P ADP | SA7255P | QNN | 34.837 ms | 0 - 9 MB | INT8 | NPU | Use Export Script |
55
- | ResNeXt101Quantized | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 2.783 ms | 0 - 28 MB | INT8 | NPU | [ResNeXt101Quantized.tflite](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.tflite) |
56
  | ResNeXt101Quantized | SA8255 (Proxy) | SA8255P Proxy | QNN | 3.021 ms | 0 - 3 MB | INT8 | NPU | Use Export Script |
57
- | ResNeXt101Quantized | SA8295P ADP | SA8295P | TFLITE | 4.012 ms | 0 - 90 MB | INT8 | NPU | [ResNeXt101Quantized.tflite](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.tflite) |
58
- | ResNeXt101Quantized | SA8295P ADP | SA8295P | QNN | 4.274 ms | 0 - 14 MB | INT8 | NPU | Use Export Script |
59
- | ResNeXt101Quantized | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 2.766 ms | 0 - 29 MB | INT8 | NPU | [ResNeXt101Quantized.tflite](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.tflite) |
60
- | ResNeXt101Quantized | SA8650 (Proxy) | SA8650P Proxy | QNN | 3.003 ms | 0 - 3 MB | INT8 | NPU | Use Export Script |
61
- | ResNeXt101Quantized | SA8775P ADP | SA8775P | TFLITE | 3.736 ms | 0 - 87 MB | INT8 | NPU | [ResNeXt101Quantized.tflite](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.tflite) |
62
- | ResNeXt101Quantized | SA8775P ADP | SA8775P | QNN | 4.185 ms | 0 - 10 MB | INT8 | NPU | Use Export Script |
63
- | ResNeXt101Quantized | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 3.178 ms | 0 - 96 MB | INT8 | NPU | [ResNeXt101Quantized.tflite](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.tflite) |
64
- | ResNeXt101Quantized | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 3.608 ms | 0 - 98 MB | INT8 | NPU | Use Export Script |
65
- | ResNeXt101Quantized | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 3.048 ms | 0 - 0 MB | INT8 | NPU | Use Export Script |
 
66
 
67
 
68
 
69
 
70
  ## Installation
71
 
72
- This model can be installed as a Python package via pip.
73
 
 
74
  ```bash
75
  pip install qai-hub-models
76
  ```
@@ -126,8 +127,8 @@ Profiling Results
126
  ResNeXt101Quantized
127
  Device : Samsung Galaxy S23 (13)
128
  Runtime : TFLITE
129
- Estimated inference time (ms) : 2.8
130
- Estimated peak memory usage (MB): [0, 33]
131
  Total # Ops : 150
132
  Compute Unit(s) : NPU (150 ops)
133
  ```
@@ -154,7 +155,7 @@ from qai_hub_models.models.resnext101_quantized import 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()
@@ -246,7 +247,8 @@ Explore all available models on [Qualcomm® AI Hub](https://aihub.qualcomm.com/)
246
 
247
 
248
  ## License
249
- * The license for the original implementation of ResNeXt101Quantized can be found [here](https://github.com/pytorch/vision/blob/main/LICENSE).
 
250
  * The license for the compiled assets for on-device deployment can be found [here](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/Qualcomm+AI+Hub+Proprietary+License.pdf)
251
 
252
 
 
36
 
37
  | Model | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
38
  |---|---|---|---|---|---|---|---|---|
39
+ | ResNeXt101Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 2.728 ms | 0 - 27 MB | INT8 | NPU | [ResNeXt101Quantized.tflite](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.tflite) |
40
+ | ResNeXt101Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 3.094 ms | 0 - 36 MB | INT8 | NPU | [ResNeXt101Quantized.so](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.so) |
41
+ | ResNeXt101Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 3.752 ms | 0 - 143 MB | INT8 | NPU | [ResNeXt101Quantized.onnx](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.onnx) |
42
+ | ResNeXt101Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 2.072 ms | 0 - 95 MB | INT8 | NPU | [ResNeXt101Quantized.tflite](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.tflite) |
43
+ | ResNeXt101Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 2.326 ms | 0 - 97 MB | INT8 | NPU | [ResNeXt101Quantized.so](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.so) |
44
+ | ResNeXt101Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 2.799 ms | 0 - 114 MB | INT8 | NPU | [ResNeXt101Quantized.onnx](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.onnx) |
45
+ | ResNeXt101Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 2.084 ms | 0 - 90 MB | INT8 | NPU | [ResNeXt101Quantized.tflite](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.tflite) |
46
+ | ResNeXt101Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 2.273 ms | 0 - 95 MB | INT8 | NPU | Use Export Script |
47
+ | ResNeXt101Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 2.554 ms | 0 - 112 MB | INT8 | NPU | [ResNeXt101Quantized.onnx](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.onnx) |
48
+ | ResNeXt101Quantized | RB3 Gen 2 (Proxy) | QCS6490 Proxy | TFLITE | 10.541 ms | 0 - 99 MB | INT8 | NPU | [ResNeXt101Quantized.tflite](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.tflite) |
49
+ | ResNeXt101Quantized | RB3 Gen 2 (Proxy) | QCS6490 Proxy | QNN | 14.751 ms | 0 - 12 MB | INT8 | NPU | Use Export Script |
50
+ | ResNeXt101Quantized | RB5 (Proxy) | QCS8250 Proxy | TFLITE | 135.145 ms | 0 - 281 MB | INT8 | GPU | [ResNeXt101Quantized.tflite](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.tflite) |
51
+ | ResNeXt101Quantized | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 2.782 ms | 0 - 28 MB | INT8 | NPU | [ResNeXt101Quantized.tflite](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.tflite) |
52
+ | ResNeXt101Quantized | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 3.028 ms | 0 - 3 MB | INT8 | NPU | Use Export Script |
53
+ | ResNeXt101Quantized | SA7255P ADP | SA7255P | TFLITE | 34.244 ms | 0 - 87 MB | INT8 | NPU | [ResNeXt101Quantized.tflite](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.tflite) |
54
+ | ResNeXt101Quantized | SA7255P ADP | SA7255P | QNN | 34.678 ms | 0 - 9 MB | INT8 | NPU | Use Export Script |
55
+ | ResNeXt101Quantized | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 2.73 ms | 0 - 30 MB | INT8 | NPU | [ResNeXt101Quantized.tflite](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.tflite) |
56
  | ResNeXt101Quantized | SA8255 (Proxy) | SA8255P Proxy | QNN | 3.021 ms | 0 - 3 MB | INT8 | NPU | Use Export Script |
57
+ | ResNeXt101Quantized | SA8295P ADP | SA8295P | TFLITE | 4.026 ms | 0 - 89 MB | INT8 | NPU | [ResNeXt101Quantized.tflite](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.tflite) |
58
+ | ResNeXt101Quantized | SA8295P ADP | SA8295P | QNN | 4.453 ms | 0 - 15 MB | INT8 | NPU | Use Export Script |
59
+ | ResNeXt101Quantized | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 2.782 ms | 0 - 29 MB | INT8 | NPU | [ResNeXt101Quantized.tflite](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.tflite) |
60
+ | ResNeXt101Quantized | SA8650 (Proxy) | SA8650P Proxy | QNN | 2.961 ms | 0 - 3 MB | INT8 | NPU | Use Export Script |
61
+ | ResNeXt101Quantized | SA8775P ADP | SA8775P | TFLITE | 3.733 ms | 0 - 87 MB | INT8 | NPU | [ResNeXt101Quantized.tflite](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.tflite) |
62
+ | ResNeXt101Quantized | SA8775P ADP | SA8775P | QNN | 4.23 ms | 0 - 10 MB | INT8 | NPU | Use Export Script |
63
+ | ResNeXt101Quantized | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 3.448 ms | 0 - 96 MB | INT8 | NPU | [ResNeXt101Quantized.tflite](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.tflite) |
64
+ | ResNeXt101Quantized | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 3.565 ms | 0 - 94 MB | INT8 | NPU | Use Export Script |
65
+ | ResNeXt101Quantized | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 3.054 ms | 0 - 0 MB | INT8 | NPU | Use Export Script |
66
+ | ResNeXt101Quantized | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 4.071 ms | 90 - 90 MB | INT8 | NPU | [ResNeXt101Quantized.onnx](https://huggingface.co/qualcomm/ResNeXt101Quantized/blob/main/ResNeXt101Quantized.onnx) |
67
 
68
 
69
 
70
 
71
  ## Installation
72
 
 
73
 
74
+ Install the package via pip:
75
  ```bash
76
  pip install qai-hub-models
77
  ```
 
127
  ResNeXt101Quantized
128
  Device : Samsung Galaxy S23 (13)
129
  Runtime : TFLITE
130
+ Estimated inference time (ms) : 2.7
131
+ Estimated peak memory usage (MB): [0, 27]
132
  Total # Ops : 150
133
  Compute Unit(s) : NPU (150 ops)
134
  ```
 
155
  torch_model = Model.from_pretrained()
156
 
157
  # Device
158
+ device = hub.Device("Samsung Galaxy S24")
159
 
160
  # Trace model
161
  input_shape = torch_model.get_input_spec()
 
247
 
248
 
249
  ## License
250
+ * The license for the original implementation of ResNeXt101Quantized can be found
251
+ [here](https://github.com/pytorch/vision/blob/main/LICENSE).
252
  * The license for the compiled assets for on-device deployment can be found [here](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/Qualcomm+AI+Hub+Proprietary+License.pdf)
253
 
254