v0.42.0
Browse filesSee https://github.com/quic/ai-hub-models/releases/v0.42.0 for changelog.
- MobileNet-v3-Small_float.dlc +2 -2
- MobileNet-v3-Small_float.onnx.zip +1 -1
- README.md +33 -33
MobileNet-v3-Small_float.dlc
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f1f5f73ef0af5480c58eeb3dd156b9f031c0b3b0919f090fbe74bd22b5c9d286
|
| 3 |
+
size 10303388
|
MobileNet-v3-Small_float.onnx.zip
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 9443857
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:011bb623cc4e850f9aec9269f95eb07f8b359d3d236458d54852a8438a5b4c72
|
| 3 |
size 9443857
|
README.md
CHANGED
|
@@ -37,36 +37,36 @@ More details on model performance across various devices, can be found
|
|
| 37 |
|
| 38 |
| Model | Precision | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit | Target Model
|
| 39 |
|---|---|---|---|---|---|---|---|---|
|
| 40 |
-
| MobileNet-v3-Small | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 2.
|
| 41 |
-
| MobileNet-v3-Small | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC | 1.
|
| 42 |
-
| MobileNet-v3-Small | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 1.
|
| 43 |
-
| MobileNet-v3-Small | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 1.
|
| 44 |
-
| MobileNet-v3-Small | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 0.
|
| 45 |
-
| MobileNet-v3-Small | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 0.
|
| 46 |
-
| MobileNet-v3-Small | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | ONNX | 0.734 ms | 0 -
|
| 47 |
-
| MobileNet-v3-Small | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 1.
|
| 48 |
-
| MobileNet-v3-Small | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC | 1.
|
| 49 |
-
| MobileNet-v3-Small | float | SA7255P ADP | Qualcomm® SA7255P | TFLITE | 2.
|
| 50 |
-
| MobileNet-v3-Small | float | SA7255P ADP | Qualcomm® SA7255P | QNN_DLC | 1.
|
| 51 |
-
| MobileNet-v3-Small | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | TFLITE | 0.
|
| 52 |
-
| MobileNet-v3-Small | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN_DLC | 0.
|
| 53 |
-
| MobileNet-v3-Small | float | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 1.
|
| 54 |
-
| MobileNet-v3-Small | float | SA8295P ADP | Qualcomm® SA8295P | QNN_DLC | 1.
|
| 55 |
-
| MobileNet-v3-Small | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | TFLITE | 0.
|
| 56 |
-
| MobileNet-v3-Small | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN_DLC | 0.
|
| 57 |
-
| MobileNet-v3-Small | float | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 1.
|
| 58 |
-
| MobileNet-v3-Small | float | SA8775P ADP | Qualcomm® SA8775P | QNN_DLC | 1.
|
| 59 |
-
| MobileNet-v3-Small | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 0.
|
| 60 |
-
| MobileNet-v3-Small | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 0.
|
| 61 |
-
| MobileNet-v3-Small | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 0.
|
| 62 |
| MobileNet-v3-Small | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | TFLITE | 0.414 ms | 0 - 28 MB | NPU | [MobileNet-v3-Small.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.tflite) |
|
| 63 |
-
| MobileNet-v3-Small | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | QNN_DLC | 0.389 ms | 1 -
|
| 64 |
-
| MobileNet-v3-Small | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | ONNX | 0.
|
| 65 |
-
| MobileNet-v3-Small | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | TFLITE | 0.
|
| 66 |
-
| MobileNet-v3-Small | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | QNN_DLC | 0.
|
| 67 |
-
| MobileNet-v3-Small | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | ONNX | 0.
|
| 68 |
-
| MobileNet-v3-Small | float | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_DLC | 0.
|
| 69 |
-
| MobileNet-v3-Small | float | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 0.
|
| 70 |
|
| 71 |
|
| 72 |
|
|
@@ -80,9 +80,9 @@ pip install qai-hub-models
|
|
| 80 |
```
|
| 81 |
|
| 82 |
|
| 83 |
-
## Configure Qualcomm® AI Hub to run this model on a cloud-hosted device
|
| 84 |
|
| 85 |
-
Sign-in to [Qualcomm® AI Hub](https://
|
| 86 |
Qualcomm® ID. Once signed in navigate to `Account -> Settings -> API Token`.
|
| 87 |
|
| 88 |
With this API token, you can configure your client to run models on the cloud
|
|
@@ -90,7 +90,7 @@ hosted devices.
|
|
| 90 |
```bash
|
| 91 |
qai-hub configure --api_token API_TOKEN
|
| 92 |
```
|
| 93 |
-
Navigate to [docs](https://
|
| 94 |
|
| 95 |
|
| 96 |
|
|
@@ -201,7 +201,7 @@ With the output of the model, you can compute like PSNR, relative errors or
|
|
| 201 |
spot check the output with expected output.
|
| 202 |
|
| 203 |
**Note**: This on-device profiling and inference requires access to Qualcomm®
|
| 204 |
-
AI Hub. [Sign up for access](https://myaccount.qualcomm.com/signup).
|
| 205 |
|
| 206 |
|
| 207 |
|
|
|
|
| 37 |
|
| 38 |
| Model | Precision | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit | Target Model
|
| 39 |
|---|---|---|---|---|---|---|---|---|
|
| 40 |
+
| MobileNet-v3-Small | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 2.077 ms | 0 - 22 MB | NPU | [MobileNet-v3-Small.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.tflite) |
|
| 41 |
+
| MobileNet-v3-Small | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC | 1.995 ms | 1 - 23 MB | NPU | [MobileNet-v3-Small.dlc](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.dlc) |
|
| 42 |
+
| MobileNet-v3-Small | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 1.012 ms | 0 - 33 MB | NPU | [MobileNet-v3-Small.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.tflite) |
|
| 43 |
+
| MobileNet-v3-Small | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 1.415 ms | 1 - 36 MB | NPU | [MobileNet-v3-Small.dlc](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.dlc) |
|
| 44 |
+
| MobileNet-v3-Small | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 0.783 ms | 0 - 50 MB | NPU | [MobileNet-v3-Small.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.tflite) |
|
| 45 |
+
| MobileNet-v3-Small | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 0.763 ms | 1 - 47 MB | NPU | [MobileNet-v3-Small.dlc](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.dlc) |
|
| 46 |
+
| MobileNet-v3-Small | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | ONNX | 0.734 ms | 0 - 54 MB | NPU | [MobileNet-v3-Small.onnx.zip](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.onnx.zip) |
|
| 47 |
+
| MobileNet-v3-Small | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 1.1 ms | 0 - 23 MB | NPU | [MobileNet-v3-Small.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.tflite) |
|
| 48 |
+
| MobileNet-v3-Small | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC | 1.057 ms | 1 - 23 MB | NPU | [MobileNet-v3-Small.dlc](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.dlc) |
|
| 49 |
+
| MobileNet-v3-Small | float | SA7255P ADP | Qualcomm® SA7255P | TFLITE | 2.077 ms | 0 - 22 MB | NPU | [MobileNet-v3-Small.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.tflite) |
|
| 50 |
+
| MobileNet-v3-Small | float | SA7255P ADP | Qualcomm® SA7255P | QNN_DLC | 1.995 ms | 1 - 23 MB | NPU | [MobileNet-v3-Small.dlc](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.dlc) |
|
| 51 |
+
| MobileNet-v3-Small | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | TFLITE | 0.786 ms | 0 - 51 MB | NPU | [MobileNet-v3-Small.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.tflite) |
|
| 52 |
+
| MobileNet-v3-Small | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN_DLC | 0.761 ms | 1 - 7 MB | NPU | [MobileNet-v3-Small.dlc](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.dlc) |
|
| 53 |
+
| MobileNet-v3-Small | float | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 1.405 ms | 0 - 30 MB | NPU | [MobileNet-v3-Small.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.tflite) |
|
| 54 |
+
| MobileNet-v3-Small | float | SA8295P ADP | Qualcomm® SA8295P | QNN_DLC | 1.385 ms | 0 - 29 MB | NPU | [MobileNet-v3-Small.dlc](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.dlc) |
|
| 55 |
+
| MobileNet-v3-Small | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | TFLITE | 0.788 ms | 0 - 50 MB | NPU | [MobileNet-v3-Small.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.tflite) |
|
| 56 |
+
| MobileNet-v3-Small | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN_DLC | 0.762 ms | 0 - 48 MB | NPU | [MobileNet-v3-Small.dlc](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.dlc) |
|
| 57 |
+
| MobileNet-v3-Small | float | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 1.1 ms | 0 - 23 MB | NPU | [MobileNet-v3-Small.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.tflite) |
|
| 58 |
+
| MobileNet-v3-Small | float | SA8775P ADP | Qualcomm® SA8775P | QNN_DLC | 1.057 ms | 1 - 23 MB | NPU | [MobileNet-v3-Small.dlc](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.dlc) |
|
| 59 |
+
| MobileNet-v3-Small | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 0.513 ms | 0 - 37 MB | NPU | [MobileNet-v3-Small.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.tflite) |
|
| 60 |
+
| MobileNet-v3-Small | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 0.509 ms | 1 - 30 MB | NPU | [MobileNet-v3-Small.dlc](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.dlc) |
|
| 61 |
+
| MobileNet-v3-Small | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 0.451 ms | 0 - 36 MB | NPU | [MobileNet-v3-Small.onnx.zip](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.onnx.zip) |
|
| 62 |
| MobileNet-v3-Small | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | TFLITE | 0.414 ms | 0 - 28 MB | NPU | [MobileNet-v3-Small.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.tflite) |
|
| 63 |
+
| MobileNet-v3-Small | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | QNN_DLC | 0.389 ms | 1 - 31 MB | NPU | [MobileNet-v3-Small.dlc](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.dlc) |
|
| 64 |
+
| MobileNet-v3-Small | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | ONNX | 0.385 ms | 0 - 24 MB | NPU | [MobileNet-v3-Small.onnx.zip](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.onnx.zip) |
|
| 65 |
+
| MobileNet-v3-Small | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | TFLITE | 0.354 ms | 0 - 29 MB | NPU | [MobileNet-v3-Small.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.tflite) |
|
| 66 |
+
| MobileNet-v3-Small | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | QNN_DLC | 0.323 ms | 0 - 26 MB | NPU | [MobileNet-v3-Small.dlc](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.dlc) |
|
| 67 |
+
| MobileNet-v3-Small | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | ONNX | 0.36 ms | 1 - 28 MB | NPU | [MobileNet-v3-Small.onnx.zip](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.onnx.zip) |
|
| 68 |
+
| MobileNet-v3-Small | float | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_DLC | 0.905 ms | 47 - 47 MB | NPU | [MobileNet-v3-Small.dlc](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.dlc) |
|
| 69 |
+
| MobileNet-v3-Small | float | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 0.665 ms | 5 - 5 MB | NPU | [MobileNet-v3-Small.onnx.zip](https://huggingface.co/qualcomm/MobileNet-v3-Small/blob/main/MobileNet-v3-Small.onnx.zip) |
|
| 70 |
|
| 71 |
|
| 72 |
|
|
|
|
| 80 |
```
|
| 81 |
|
| 82 |
|
| 83 |
+
## Configure Qualcomm® AI Hub Workbench to run this model on a cloud-hosted device
|
| 84 |
|
| 85 |
+
Sign-in to [Qualcomm® AI Hub Workbench](https://workbench.aihub.qualcomm.com/) with your
|
| 86 |
Qualcomm® ID. Once signed in navigate to `Account -> Settings -> API Token`.
|
| 87 |
|
| 88 |
With this API token, you can configure your client to run models on the cloud
|
|
|
|
| 90 |
```bash
|
| 91 |
qai-hub configure --api_token API_TOKEN
|
| 92 |
```
|
| 93 |
+
Navigate to [docs](https://workbench.aihub.qualcomm.com/docs/) for more information.
|
| 94 |
|
| 95 |
|
| 96 |
|
|
|
|
| 201 |
spot check the output with expected output.
|
| 202 |
|
| 203 |
**Note**: This on-device profiling and inference requires access to Qualcomm®
|
| 204 |
+
AI Hub Workbench. [Sign up for access](https://myaccount.qualcomm.com/signup).
|
| 205 |
|
| 206 |
|
| 207 |
|