Hugging Face's logo

qualcomm
/
Swin-Small

Image Classification
PyTorch
TF Lite
ONNX
backbone
android
Model card Files Community
qaihm-bot commited on
Commit
4a51011
·
verified ·
1 Parent(s): 901c933

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +31 -30
README.md CHANGED
@@ -35,39 +35,39 @@ 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
- | Swin-Small | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 19.265 ms | 0 - 41 MB | FP16 | NPU | [Swin-Small.tflite](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.tflite) |
39
- | Swin-Small | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 19.301 ms | 0 - 41 MB | FP16 | NPU | [Swin-Small.so](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.so) |
40
- | Swin-Small | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 31.628 ms | 0 - 305 MB | FP16 | NPU | [Swin-Small.onnx](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.onnx) |
41
- | Swin-Small | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 13.265 ms | 0 - 159 MB | FP16 | NPU | [Swin-Small.tflite](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.tflite) |
42
- | Swin-Small | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 13.055 ms | 1 - 159 MB | FP16 | NPU | [Swin-Small.so](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.so) |
43
- | Swin-Small | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 21.208 ms | 1 - 183 MB | FP16 | NPU | [Swin-Small.onnx](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.onnx) |
44
- | Swin-Small | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 12.088 ms | 0 - 162 MB | FP16 | NPU | [Swin-Small.tflite](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.tflite) |
45
- | Swin-Small | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 11.834 ms | 0 - 164 MB | FP16 | NPU | Use Export Script |
46
- | Swin-Small | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 18.991 ms | 1 - 195 MB | FP16 | NPU | [Swin-Small.onnx](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.onnx) |
47
- | Swin-Small | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 19.233 ms | 0 - 38 MB | FP16 | NPU | [Swin-Small.tflite](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.tflite) |
48
- | Swin-Small | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 17.704 ms | 1 - 3 MB | FP16 | NPU | Use Export Script |
49
- | Swin-Small | SA7255P ADP | SA7255P | TFLITE | 181.603 ms | 0 - 165 MB | FP16 | NPU | [Swin-Small.tflite](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.tflite) |
50
- | Swin-Small | SA7255P ADP | SA7255P | QNN | 177.664 ms | 1 - 10 MB | FP16 | NPU | Use Export Script |
51
- | Swin-Small | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 19.491 ms | 0 - 43 MB | FP16 | NPU | [Swin-Small.tflite](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.tflite) |
52
- | Swin-Small | SA8255 (Proxy) | SA8255P Proxy | QNN | 17.635 ms | 2 - 4 MB | FP16 | NPU | Use Export Script |
53
- | Swin-Small | SA8295P ADP | SA8295P | TFLITE | 27.438 ms | 0 - 155 MB | FP16 | NPU | [Swin-Small.tflite](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.tflite) |
54
- | Swin-Small | SA8295P ADP | SA8295P | QNN | 25.368 ms | 1 - 15 MB | FP16 | NPU | Use Export Script |
55
- | Swin-Small | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 19.371 ms | 0 - 38 MB | FP16 | NPU | [Swin-Small.tflite](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.tflite) |
56
- | Swin-Small | SA8650 (Proxy) | SA8650P Proxy | QNN | 17.55 ms | 1 - 3 MB | FP16 | NPU | Use Export Script |
57
- | Swin-Small | SA8775P ADP | SA8775P | TFLITE | 26.043 ms | 0 - 165 MB | FP16 | NPU | [Swin-Small.tflite](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.tflite) |
58
- | Swin-Small | SA8775P ADP | SA8775P | QNN | 23.984 ms | 1 - 11 MB | FP16 | NPU | Use Export Script |
59
- | Swin-Small | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 24.434 ms | 0 - 154 MB | FP16 | NPU | [Swin-Small.tflite](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.tflite) |
60
- | Swin-Small | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 24.364 ms | 0 - 155 MB | FP16 | NPU | Use Export Script |
61
- | Swin-Small | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 18.423 ms | 1 - 1 MB | FP16 | NPU | Use Export Script |
62
- | Swin-Small | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 38.128 ms | 100 - 100 MB | FP16 | NPU | [Swin-Small.onnx](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.onnx) |
63
 
64
 
65
 
66
 
67
  ## Installation
68
 
69
- This model can be installed as a Python package via pip.
70
 
 
71
  ```bash
72
  pip install qai-hub-models
73
  ```
@@ -123,8 +123,8 @@ Profiling Results
123
  Swin-Small
124
  Device : Samsung Galaxy S23 (13)
125
  Runtime : TFLITE
126
- Estimated inference time (ms) : 19.3
127
- Estimated peak memory usage (MB): [0, 41]
128
  Total # Ops : 1563
129
  Compute Unit(s) : NPU (1563 ops)
130
  ```
@@ -151,7 +151,7 @@ from qai_hub_models.models.swin_small import 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()
@@ -243,7 +243,8 @@ Explore all available models on [Qualcomm® AI Hub](https://aihub.qualcomm.com/)
243
 
244
 
245
  ## License
246
- * The license for the original implementation of Swin-Small can be found [here](https://github.com/pytorch/vision/blob/main/LICENSE).
 
247
  * 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)
248
 
249
 
 
35
 
36
  | Model | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
37
  |---|---|---|---|---|---|---|---|---|
38
+ | Swin-Small | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 19.423 ms | 0 - 35 MB | FP16 | NPU | [Swin-Small.tflite](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.tflite) |
39
+ | Swin-Small | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 19.238 ms | 0 - 39 MB | FP16 | NPU | [Swin-Small.so](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.so) |
40
+ | Swin-Small | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 31.292 ms | 0 - 243 MB | FP16 | NPU | [Swin-Small.onnx](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.onnx) |
41
+ | Swin-Small | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 13.122 ms | 0 - 157 MB | FP16 | NPU | [Swin-Small.tflite](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.tflite) |
42
+ | Swin-Small | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 12.985 ms | 1 - 161 MB | FP16 | NPU | [Swin-Small.so](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.so) |
43
+ | Swin-Small | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 21.265 ms | 1 - 181 MB | FP16 | NPU | [Swin-Small.onnx](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.onnx) |
44
+ | Swin-Small | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 12.108 ms | 0 - 163 MB | FP16 | NPU | [Swin-Small.tflite](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.tflite) |
45
+ | Swin-Small | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 9.94 ms | 1 - 166 MB | FP16 | NPU | Use Export Script |
46
+ | Swin-Small | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 18.965 ms | 1 - 189 MB | FP16 | NPU | [Swin-Small.onnx](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.onnx) |
47
+ | Swin-Small | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 19.204 ms | 0 - 39 MB | FP16 | NPU | [Swin-Small.tflite](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.tflite) |
48
+ | Swin-Small | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 17.679 ms | 1 - 3 MB | FP16 | NPU | Use Export Script |
49
+ | Swin-Small | SA7255P ADP | SA7255P | TFLITE | 181.551 ms | 0 - 165 MB | FP16 | NPU | [Swin-Small.tflite](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.tflite) |
50
+ | Swin-Small | SA7255P ADP | SA7255P | QNN | 177.493 ms | 1 - 9 MB | FP16 | NPU | Use Export Script |
51
+ | Swin-Small | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 19.344 ms | 0 - 37 MB | FP16 | NPU | [Swin-Small.tflite](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.tflite) |
52
+ | Swin-Small | SA8255 (Proxy) | SA8255P Proxy | QNN | 17.69 ms | 1 - 3 MB | FP16 | NPU | Use Export Script |
53
+ | Swin-Small | SA8295P ADP | SA8295P | TFLITE | 27.385 ms | 0 - 155 MB | FP16 | NPU | [Swin-Small.tflite](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.tflite) |
54
+ | Swin-Small | SA8295P ADP | SA8295P | QNN | 25.297 ms | 1 - 15 MB | FP16 | NPU | Use Export Script |
55
+ | Swin-Small | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 19.363 ms | 0 - 39 MB | FP16 | NPU | [Swin-Small.tflite](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.tflite) |
56
+ | Swin-Small | SA8650 (Proxy) | SA8650P Proxy | QNN | 17.706 ms | 1 - 3 MB | FP16 | NPU | Use Export Script |
57
+ | Swin-Small | SA8775P ADP | SA8775P | TFLITE | 26.069 ms | 0 - 165 MB | FP16 | NPU | [Swin-Small.tflite](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.tflite) |
58
+ | Swin-Small | SA8775P ADP | SA8775P | QNN | 23.912 ms | 1 - 11 MB | FP16 | NPU | Use Export Script |
59
+ | Swin-Small | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 24.606 ms | 0 - 154 MB | FP16 | NPU | [Swin-Small.tflite](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.tflite) |
60
+ | Swin-Small | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 24.656 ms | 0 - 155 MB | FP16 | NPU | Use Export Script |
61
+ | Swin-Small | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 18.425 ms | 1 - 1 MB | FP16 | NPU | Use Export Script |
62
+ | Swin-Small | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 37.968 ms | 101 - 101 MB | FP16 | NPU | [Swin-Small.onnx](https://huggingface.co/qualcomm/Swin-Small/blob/main/Swin-Small.onnx) |
63
 
64
 
65
 
66
 
67
  ## Installation
68
 
 
69
 
70
+ Install the package via pip:
71
  ```bash
72
  pip install qai-hub-models
73
  ```
 
123
  Swin-Small
124
  Device : Samsung Galaxy S23 (13)
125
  Runtime : TFLITE
126
+ Estimated inference time (ms) : 19.4
127
+ Estimated peak memory usage (MB): [0, 35]
128
  Total # Ops : 1563
129
  Compute Unit(s) : NPU (1563 ops)
130
  ```
 
151
  torch_model = Model.from_pretrained()
152
 
153
  # Device
154
+ device = hub.Device("Samsung Galaxy S24")
155
 
156
  # Trace model
157
  input_shape = torch_model.get_input_spec()
 
243
 
244
 
245
  ## License
246
+ * The license for the original implementation of Swin-Small can be found
247
+ [here](https://github.com/pytorch/vision/blob/main/LICENSE).
248
  * 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)
249
 
250