Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
@@ -35,33 +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 |
-
| ResNet-2Plus1D | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 116.
|
39 |
-
| ResNet-2Plus1D | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 14.
|
40 |
-
| ResNet-2Plus1D | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE |
|
41 |
-
| ResNet-2Plus1D | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 10.
|
42 |
-
| ResNet-2Plus1D | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE |
|
43 |
-
| ResNet-2Plus1D | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX |
|
44 |
-
| ResNet-2Plus1D | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 116.
|
45 |
-
| ResNet-2Plus1D | SA7255P ADP | SA7255P | TFLITE |
|
46 |
-
| ResNet-2Plus1D | SA8255 (Proxy) | SA8255P Proxy | TFLITE |
|
47 |
-
| ResNet-2Plus1D | SA8295P ADP | SA8295P | TFLITE |
|
48 |
-
| ResNet-2Plus1D | SA8650 (Proxy) | SA8650P Proxy | TFLITE |
|
49 |
-
| ResNet-2Plus1D | SA8775P ADP | SA8775P | TFLITE | 156.
|
50 |
-
| ResNet-2Plus1D | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE |
|
|
|
51 |
|
52 |
|
53 |
|
54 |
|
55 |
## Installation
|
56 |
|
57 |
-
This model can be installed as a Python package via pip.
|
58 |
|
|
|
59 |
```bash
|
60 |
-
pip install "qai-hub-models[
|
61 |
```
|
62 |
|
63 |
|
64 |
-
|
65 |
## Configure Qualcomm® AI Hub to run this model on a cloud-hosted device
|
66 |
|
67 |
Sign-in to [Qualcomm® AI Hub](https://app.aihub.qualcomm.com/) with your
|
@@ -112,7 +112,7 @@ Profiling Results
|
|
112 |
ResNet-2Plus1D
|
113 |
Device : Samsung Galaxy S23 (13)
|
114 |
Runtime : TFLITE
|
115 |
-
Estimated inference time (ms) : 116.
|
116 |
Estimated peak memory usage (MB): [4, 601]
|
117 |
Total # Ops : 92
|
118 |
Compute Unit(s) : NPU (84 ops) CPU (8 ops)
|
@@ -140,7 +140,7 @@ from qai_hub_models.models.resnet_2plus1d import Model
|
|
140 |
torch_model = Model.from_pretrained()
|
141 |
|
142 |
# Device
|
143 |
-
device = hub.Device("Samsung Galaxy
|
144 |
|
145 |
# Trace model
|
146 |
input_shape = torch_model.get_input_spec()
|
@@ -218,7 +218,8 @@ Explore all available models on [Qualcomm® AI Hub](https://aihub.qualcomm.com/)
|
|
218 |
|
219 |
|
220 |
## License
|
221 |
-
* The license for the original implementation of ResNet-2Plus1D can be found
|
|
|
222 |
* 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)
|
223 |
|
224 |
|
|
|
35 |
|
36 |
| Model | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
|
37 |
|---|---|---|---|---|---|---|---|---|
|
38 |
+
| ResNet-2Plus1D | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 116.022 ms | 4 - 601 MB | FP16 | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.tflite) |
|
39 |
+
| ResNet-2Plus1D | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 14.287 ms | 0 - 147 MB | FP16 | NPU | [ResNet-2Plus1D.onnx](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.onnx) |
|
40 |
+
| ResNet-2Plus1D | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 90.026 ms | 27 - 66 MB | FP16 | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.tflite) |
|
41 |
+
| ResNet-2Plus1D | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 10.487 ms | 2 - 62 MB | FP16 | NPU | [ResNet-2Plus1D.onnx](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.onnx) |
|
42 |
+
| ResNet-2Plus1D | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 87.279 ms | 27 - 67 MB | FP16 | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.tflite) |
|
43 |
+
| ResNet-2Plus1D | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 8.446 ms | 2 - 62 MB | FP16 | NPU | [ResNet-2Plus1D.onnx](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.onnx) |
|
44 |
+
| ResNet-2Plus1D | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 116.79 ms | 4 - 605 MB | FP16 | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.tflite) |
|
45 |
+
| ResNet-2Plus1D | SA7255P ADP | SA7255P | TFLITE | 828.253 ms | 28 - 63 MB | FP16 | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.tflite) |
|
46 |
+
| ResNet-2Plus1D | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 117.896 ms | 4 - 604 MB | FP16 | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.tflite) |
|
47 |
+
| ResNet-2Plus1D | SA8295P ADP | SA8295P | TFLITE | 142.683 ms | 28 - 65 MB | FP16 | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.tflite) |
|
48 |
+
| ResNet-2Plus1D | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 116.463 ms | 0 - 601 MB | FP16 | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.tflite) |
|
49 |
+
| ResNet-2Plus1D | SA8775P ADP | SA8775P | TFLITE | 156.295 ms | 28 - 64 MB | FP16 | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.tflite) |
|
50 |
+
| ResNet-2Plus1D | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 129.52 ms | 28 - 71 MB | FP16 | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.tflite) |
|
51 |
+
| ResNet-2Plus1D | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 14.848 ms | 62 - 62 MB | FP16 | NPU | [ResNet-2Plus1D.onnx](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.onnx) |
|
52 |
|
53 |
|
54 |
|
55 |
|
56 |
## Installation
|
57 |
|
|
|
58 |
|
59 |
+
Install the package via pip:
|
60 |
```bash
|
61 |
+
pip install "qai-hub-models[resnet-2plus1d]"
|
62 |
```
|
63 |
|
64 |
|
|
|
65 |
## Configure Qualcomm® AI Hub to run this model on a cloud-hosted device
|
66 |
|
67 |
Sign-in to [Qualcomm® AI Hub](https://app.aihub.qualcomm.com/) with your
|
|
|
112 |
ResNet-2Plus1D
|
113 |
Device : Samsung Galaxy S23 (13)
|
114 |
Runtime : TFLITE
|
115 |
+
Estimated inference time (ms) : 116.0
|
116 |
Estimated peak memory usage (MB): [4, 601]
|
117 |
Total # Ops : 92
|
118 |
Compute Unit(s) : NPU (84 ops) CPU (8 ops)
|
|
|
140 |
torch_model = Model.from_pretrained()
|
141 |
|
142 |
# Device
|
143 |
+
device = hub.Device("Samsung Galaxy S24")
|
144 |
|
145 |
# Trace model
|
146 |
input_shape = torch_model.get_input_spec()
|
|
|
218 |
|
219 |
|
220 |
## License
|
221 |
+
* The license for the original implementation of ResNet-2Plus1D can be found
|
222 |
+
[here](https://github.com/pytorch/vision/blob/main/LICENSE).
|
223 |
* 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)
|
224 |
|
225 |
|