Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
@@ -32,8 +32,8 @@ More details on model performance across various devices, can be found
|
|
32 |
|
33 |
| Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
|
34 |
| ---|---|---|---|---|---|---|---|
|
35 |
-
| Samsung Galaxy S23 Ultra (Android 13) | Snapdragon® 8 Gen 2 | TFLite | 8.
|
36 |
-
| Samsung Galaxy S23 Ultra (Android 13) | Snapdragon® 8 Gen 2 | QNN Model Library | 7.
|
37 |
|
38 |
|
39 |
## Installation
|
@@ -93,16 +93,16 @@ python -m qai_hub_models.models.fcn_resnet50.export
|
|
93 |
```
|
94 |
Profile Job summary of FCN_ResNet50
|
95 |
--------------------------------------------------
|
96 |
-
Device: Samsung Galaxy
|
97 |
-
Estimated Inference Time:
|
98 |
-
Estimated Peak Memory Range: 4.
|
99 |
Compute Units: NPU (84) | Total (84)
|
100 |
|
101 |
Profile Job summary of FCN_ResNet50
|
102 |
--------------------------------------------------
|
103 |
-
Device: Samsung Galaxy
|
104 |
-
Estimated Inference Time:
|
105 |
-
Estimated Peak Memory Range: 0.
|
106 |
Compute Units: NPU (126) | Total (126)
|
107 |
|
108 |
|
@@ -222,7 +222,7 @@ Explore all available models on [Qualcomm® AI Hub](https://aihub.qualcomm.com/)
|
|
222 |
## License
|
223 |
- The license for the original implementation of FCN_ResNet50 can be found
|
224 |
[here](https://github.com/pytorch/vision/blob/main/LICENSE).
|
225 |
-
- The license for the compiled assets for on-device deployment can be found [here](
|
226 |
|
227 |
## References
|
228 |
* [Fully Convolutional Networks for Semantic Segmentation](https://arxiv.org/abs/1411.4038)
|
|
|
32 |
|
33 |
| Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
|
34 |
| ---|---|---|---|---|---|---|---|
|
35 |
+
| Samsung Galaxy S23 Ultra (Android 13) | Snapdragon® 8 Gen 2 | TFLite | 8.55 ms | 4 - 6 MB | FP16 | NPU | [FCN_ResNet50.tflite](https://huggingface.co/qualcomm/FCN_ResNet50/blob/main/FCN_ResNet50.tflite)
|
36 |
+
| Samsung Galaxy S23 Ultra (Android 13) | Snapdragon® 8 Gen 2 | QNN Model Library | 7.881 ms | 0 - 13 MB | FP16 | NPU | [FCN_ResNet50.so](https://huggingface.co/qualcomm/FCN_ResNet50/blob/main/FCN_ResNet50.so)
|
37 |
|
38 |
|
39 |
## Installation
|
|
|
93 |
```
|
94 |
Profile Job summary of FCN_ResNet50
|
95 |
--------------------------------------------------
|
96 |
+
Device: Samsung Galaxy S24 (14)
|
97 |
+
Estimated Inference Time: 6.41 ms
|
98 |
+
Estimated Peak Memory Range: 4.05-72.84 MB
|
99 |
Compute Units: NPU (84) | Total (84)
|
100 |
|
101 |
Profile Job summary of FCN_ResNet50
|
102 |
--------------------------------------------------
|
103 |
+
Device: Samsung Galaxy S24 (14)
|
104 |
+
Estimated Inference Time: 5.85 ms
|
105 |
+
Estimated Peak Memory Range: 0.61-53.34 MB
|
106 |
Compute Units: NPU (126) | Total (126)
|
107 |
|
108 |
|
|
|
222 |
## License
|
223 |
- The license for the original implementation of FCN_ResNet50 can be found
|
224 |
[here](https://github.com/pytorch/vision/blob/main/LICENSE).
|
225 |
+
- The license for the compiled assets for on-device deployment can be found [here]({deploy_license_url})
|
226 |
|
227 |
## References
|
228 |
* [Fully Convolutional Networks for Semantic Segmentation](https://arxiv.org/abs/1411.4038)
|