Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
@@ -37,7 +37,8 @@ More details on model performance across various devices, can be found
|
|
37 |
|
38 |
| Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
|
39 |
| ---|---|---|---|---|---|---|---|
|
40 |
-
| Samsung Galaxy S23 Ultra (Android 13) | Snapdragon® 8 Gen 2 | TFLite |
|
|
|
41 |
|
42 |
|
43 |
## Installation
|
@@ -97,10 +98,17 @@ python -m qai_hub_models.models.resnet101_quantized.export
|
|
97 |
```
|
98 |
Profile Job summary of ResNet101Quantized
|
99 |
--------------------------------------------------
|
100 |
-
Device: Samsung Galaxy
|
101 |
-
Estimated Inference Time:
|
102 |
-
Estimated Peak Memory Range: 0.
|
103 |
-
Compute Units: NPU (
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
104 |
|
105 |
|
106 |
```
|
@@ -219,7 +227,7 @@ Explore all available models on [Qualcomm® AI Hub](https://aihub.qualcomm.com/)
|
|
219 |
## License
|
220 |
- The license for the original implementation of ResNet101Quantized can be found
|
221 |
[here](https://github.com/pytorch/vision/blob/main/LICENSE).
|
222 |
-
- The license for the compiled assets for on-device deployment can be found [here](
|
223 |
|
224 |
## References
|
225 |
* [Deep Residual Learning for Image Recognition](https://arxiv.org/abs/1512.03385)
|
|
|
37 |
|
38 |
| Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
|
39 |
| ---|---|---|---|---|---|---|---|
|
40 |
+
| Samsung Galaxy S23 Ultra (Android 13) | Snapdragon® 8 Gen 2 | TFLite | 1.122 ms | 0 - 2 MB | INT8 | NPU | [ResNet101Quantized.tflite](https://huggingface.co/qualcomm/ResNet101Quantized/blob/main/ResNet101Quantized.tflite)
|
41 |
+
| Samsung Galaxy S23 Ultra (Android 13) | Snapdragon® 8 Gen 2 | QNN Model Library | 1.101 ms | 0 - 188 MB | INT8 | NPU | [ResNet101Quantized.so](https://huggingface.co/qualcomm/ResNet101Quantized/blob/main/ResNet101Quantized.so)
|
42 |
|
43 |
|
44 |
## Installation
|
|
|
98 |
```
|
99 |
Profile Job summary of ResNet101Quantized
|
100 |
--------------------------------------------------
|
101 |
+
Device: Samsung Galaxy S24 (14)
|
102 |
+
Estimated Inference Time: 0.84 ms
|
103 |
+
Estimated Peak Memory Range: 0.01-87.01 MB
|
104 |
+
Compute Units: NPU (146) | Total (146)
|
105 |
+
|
106 |
+
Profile Job summary of ResNet101Quantized
|
107 |
+
--------------------------------------------------
|
108 |
+
Device: Samsung Galaxy S24 (14)
|
109 |
+
Estimated Inference Time: 0.83 ms
|
110 |
+
Estimated Peak Memory Range: 0.16-51.47 MB
|
111 |
+
Compute Units: NPU (144) | Total (144)
|
112 |
|
113 |
|
114 |
```
|
|
|
227 |
## License
|
228 |
- The license for the original implementation of ResNet101Quantized can be found
|
229 |
[here](https://github.com/pytorch/vision/blob/main/LICENSE).
|
230 |
+
- The license for the compiled assets for on-device deployment can be found [here]({deploy_license_url})
|
231 |
|
232 |
## References
|
233 |
* [Deep Residual Learning for Image Recognition](https://arxiv.org/abs/1512.03385)
|