v0.53.0
Browse filesSee https://github.com/qualcomm/ai-hub-models/releases/v0.53.0 for changelog.
- README.md +79 -61
- release_assets.json +5 -5
README.md
CHANGED
|
@@ -27,9 +27,9 @@ Below are pre-exported model assets ready for deployment.
|
|
| 27 |
|
| 28 |
| Runtime | Precision | Chipset | SDK Versions | Download |
|
| 29 |
|---|---|---|---|---|
|
| 30 |
-
| ONNX | float | Universal | QAIRT 2.42, ONNX Runtime 1.24.3 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/distil_whisper/releases/v0.
|
| 31 |
-
| QNN_DLC | float | Universal | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/distil_whisper/releases/v0.
|
| 32 |
-
| TFLITE | float | Universal | | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/distil_whisper/releases/v0.
|
| 33 |
|
| 34 |
For more device-specific assets and performance metrics, visit **[Distil-Whisper on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/distil_whisper)**.
|
| 35 |
|
|
@@ -61,64 +61,82 @@ See our repository for [Distil-Whisper on GitHub](https://github.com/qualcomm/ai
|
|
| 61 |
## Performance Summary
|
| 62 |
| Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
|
| 63 |
|---|---|---|---|---|---|---
|
| 64 |
-
| decoder | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 5.
|
| 65 |
-
| decoder | ONNX | float | Snapdragon®
|
| 66 |
-
| decoder | ONNX | float | Snapdragon®
|
| 67 |
-
| decoder | ONNX | float | Snapdragon®
|
| 68 |
-
| decoder | ONNX | float |
|
| 69 |
-
| decoder | ONNX | float |
|
| 70 |
-
| decoder | ONNX | float |
|
| 71 |
-
| decoder |
|
| 72 |
-
| decoder |
|
| 73 |
-
| decoder | QNN_DLC | float | Snapdragon®
|
| 74 |
-
| decoder | QNN_DLC | float | Snapdragon® 8
|
| 75 |
-
| decoder | QNN_DLC | float |
|
| 76 |
-
| decoder | QNN_DLC | float |
|
| 77 |
-
| decoder | QNN_DLC | float |
|
| 78 |
-
| decoder | QNN_DLC | float |
|
| 79 |
-
| decoder | QNN_DLC | float | Qualcomm®
|
| 80 |
-
| decoder | QNN_DLC | float | Qualcomm®
|
| 81 |
-
| decoder | QNN_DLC | float | Qualcomm®
|
| 82 |
-
| decoder | QNN_DLC | float |
|
| 83 |
-
| decoder |
|
| 84 |
-
| decoder |
|
| 85 |
-
| decoder |
|
| 86 |
-
| decoder |
|
| 87 |
-
| decoder |
|
| 88 |
-
| decoder |
|
| 89 |
-
| decoder | TFLITE | float |
|
| 90 |
-
| decoder | TFLITE | float |
|
| 91 |
-
| decoder | TFLITE | float |
|
| 92 |
-
| decoder | TFLITE | float |
|
| 93 |
-
|
|
| 94 |
-
|
|
| 95 |
-
|
|
| 96 |
-
|
|
| 97 |
-
|
|
| 98 |
-
|
|
| 99 |
-
|
|
| 100 |
-
|
|
| 101 |
-
|
|
| 102 |
-
| encoder |
|
| 103 |
-
| encoder |
|
| 104 |
-
| encoder |
|
| 105 |
-
| encoder |
|
| 106 |
-
| encoder |
|
| 107 |
-
| encoder |
|
| 108 |
-
| encoder |
|
| 109 |
-
| encoder |
|
| 110 |
-
| encoder |
|
| 111 |
-
| encoder | QNN_DLC | float | Snapdragon® 8 Elite
|
| 112 |
-
| encoder |
|
| 113 |
-
| encoder |
|
| 114 |
-
| encoder |
|
| 115 |
-
| encoder |
|
| 116 |
-
| encoder |
|
| 117 |
-
| encoder |
|
| 118 |
-
| encoder |
|
| 119 |
-
| encoder |
|
| 120 |
-
| encoder |
|
| 121 |
-
| encoder |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 122 |
|
| 123 |
## License
|
| 124 |
* The license for the original implementation of Distil-Whisper can be found
|
|
|
|
| 27 |
|
| 28 |
| Runtime | Precision | Chipset | SDK Versions | Download |
|
| 29 |
|---|---|---|---|---|
|
| 30 |
+
| ONNX | float | Universal | QAIRT 2.42, ONNX Runtime 1.24.3 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/distil_whisper/releases/v0.53.0/distil_whisper-onnx-float.zip)
|
| 31 |
+
| QNN_DLC | float | Universal | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/distil_whisper/releases/v0.53.0/distil_whisper-qnn_dlc-float.zip)
|
| 32 |
+
| TFLITE | float | Universal | | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/distil_whisper/releases/v0.53.0/distil_whisper-tflite-float.zip)
|
| 33 |
|
| 34 |
For more device-specific assets and performance metrics, visit **[Distil-Whisper on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/distil_whisper)**.
|
| 35 |
|
|
|
|
| 61 |
## Performance Summary
|
| 62 |
| Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
|
| 63 |
|---|---|---|---|---|---|---
|
| 64 |
+
| decoder | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 5.546 ms | 52 - 375 MB | NPU
|
| 65 |
+
| decoder | ONNX | float | Snapdragon® 8 Elite Mobile | 7.178 ms | 16 - 472 MB | NPU
|
| 66 |
+
| decoder | ONNX | float | Snapdragon® X2 Elite | 5.083 ms | 178 - 178 MB | NPU
|
| 67 |
+
| decoder | ONNX | float | Snapdragon® X Elite | 11.41 ms | 178 - 178 MB | NPU
|
| 68 |
+
| decoder | ONNX | float | Snapdragon® X Elite | 11.41 ms | 178 - 178 MB | NPU
|
| 69 |
+
| decoder | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 8.67 ms | 52 - 426 MB | NPU
|
| 70 |
+
| decoder | ONNX | float | Qualcomm® QCS8550 (Proxy) | 11.746 ms | 0 - 184 MB | NPU
|
| 71 |
+
| decoder | ONNX | float | Qualcomm® QCS9075 | 13.188 ms | 40 - 82 MB | NPU
|
| 72 |
+
| decoder | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 7.178 ms | 16 - 472 MB | NPU
|
| 73 |
+
| decoder | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 5.92 ms | 5 - 504 MB | NPU
|
| 74 |
+
| decoder | QNN_DLC | float | Snapdragon® 8 Elite Mobile | 7.295 ms | 5 - 548 MB | NPU
|
| 75 |
+
| decoder | QNN_DLC | float | Snapdragon® X2 Elite | 6.086 ms | 40 - 40 MB | NPU
|
| 76 |
+
| decoder | QNN_DLC | float | Snapdragon® X Elite | 10.901 ms | 40 - 40 MB | NPU
|
| 77 |
+
| decoder | QNN_DLC | float | Snapdragon® X Elite | 10.901 ms | 40 - 40 MB | NPU
|
| 78 |
+
| decoder | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 8.613 ms | 0 - 601 MB | NPU
|
| 79 |
+
| decoder | QNN_DLC | float | Qualcomm® QCS8275 (Proxy) | 19.248 ms | 28 - 525 MB | NPU
|
| 80 |
+
| decoder | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 11.477 ms | 40 - 43 MB | NPU
|
| 81 |
+
| decoder | QNN_DLC | float | Qualcomm® SA8775P | 12.828 ms | 30 - 522 MB | NPU
|
| 82 |
+
| decoder | QNN_DLC | float | Qualcomm® SA8775P | 12.828 ms | 30 - 522 MB | NPU
|
| 83 |
+
| decoder | QNN_DLC | float | Qualcomm® SA8775P | 12.828 ms | 30 - 522 MB | NPU
|
| 84 |
+
| decoder | QNN_DLC | float | Qualcomm® QCS9075 | 16.542 ms | 40 - 86 MB | NPU
|
| 85 |
+
| decoder | QNN_DLC | float | Qualcomm® QCS8450 (Proxy) | 18.114 ms | 26 - 330 MB | NPU
|
| 86 |
+
| decoder | QNN_DLC | float | Qualcomm® SA7255P | 19.248 ms | 28 - 525 MB | NPU
|
| 87 |
+
| decoder | QNN_DLC | float | Qualcomm® SA8295P | 14.11 ms | 34 - 276 MB | NPU
|
| 88 |
+
| decoder | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 7.295 ms | 5 - 548 MB | NPU
|
| 89 |
+
| decoder | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 5.848 ms | 4 - 570 MB | NPU
|
| 90 |
+
| decoder | TFLITE | float | Snapdragon® 8 Elite Mobile | 7.215 ms | 4 - 572 MB | NPU
|
| 91 |
+
| decoder | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 8.619 ms | 4 - 749 MB | NPU
|
| 92 |
+
| decoder | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 19.275 ms | 4 - 537 MB | NPU
|
| 93 |
+
| decoder | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 11.713 ms | 5 - 7 MB | NPU
|
| 94 |
+
| decoder | TFLITE | float | Qualcomm® SA8775P | 13.037 ms | 5 - 537 MB | NPU
|
| 95 |
+
| decoder | TFLITE | float | Qualcomm® SA8775P | 13.037 ms | 5 - 537 MB | NPU
|
| 96 |
+
| decoder | TFLITE | float | Qualcomm® SA8775P | 13.037 ms | 5 - 537 MB | NPU
|
| 97 |
+
| decoder | TFLITE | float | Qualcomm® QCS9075 | 16.188 ms | 0 - 265 MB | NPU
|
| 98 |
+
| decoder | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 18.522 ms | 5 - 466 MB | NPU
|
| 99 |
+
| decoder | TFLITE | float | Qualcomm® SA7255P | 19.275 ms | 4 - 537 MB | NPU
|
| 100 |
+
| decoder | TFLITE | float | Qualcomm® SA8295P | 14.231 ms | 5 - 297 MB | NPU
|
| 101 |
+
| decoder | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 7.215 ms | 4 - 572 MB | NPU
|
| 102 |
+
| encoder | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 49.731 ms | 80 - 839 MB | NPU
|
| 103 |
+
| encoder | ONNX | float | Snapdragon® 8 Elite Mobile | 60.575 ms | 80 - 734 MB | NPU
|
| 104 |
+
| encoder | ONNX | float | Snapdragon® X2 Elite | 50.437 ms | 183 - 183 MB | NPU
|
| 105 |
+
| encoder | ONNX | float | Snapdragon® X Elite | 123.445 ms | 182 - 182 MB | NPU
|
| 106 |
+
| encoder | ONNX | float | Snapdragon® X Elite | 123.445 ms | 182 - 182 MB | NPU
|
| 107 |
+
| encoder | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 82.214 ms | 80 - 1239 MB | NPU
|
| 108 |
+
| encoder | ONNX | float | Qualcomm® QCS8550 (Proxy) | 118.816 ms | 0 - 202 MB | NPU
|
| 109 |
+
| encoder | ONNX | float | Qualcomm® QCS9075 | 150.604 ms | 79 - 83 MB | NPU
|
| 110 |
+
| encoder | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 60.575 ms | 80 - 734 MB | NPU
|
| 111 |
+
| encoder | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 58.551 ms | 1 - 712 MB | NPU
|
| 112 |
+
| encoder | QNN_DLC | float | Snapdragon® 8 Elite Mobile | 71.602 ms | 1 - 692 MB | NPU
|
| 113 |
+
| encoder | QNN_DLC | float | Snapdragon® X2 Elite | 60.24 ms | 1 - 1 MB | NPU
|
| 114 |
+
| encoder | QNN_DLC | float | Snapdragon® X Elite | 139.101 ms | 1 - 1 MB | NPU
|
| 115 |
+
| encoder | QNN_DLC | float | Snapdragon® X Elite | 139.101 ms | 1 - 1 MB | NPU
|
| 116 |
+
| encoder | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 97.219 ms | 0 - 962 MB | NPU
|
| 117 |
+
| encoder | QNN_DLC | float | Qualcomm® QCS8275 (Proxy) | 437.836 ms | 1 - 696 MB | NPU
|
| 118 |
+
| encoder | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 135.253 ms | 0 - 7 MB | NPU
|
| 119 |
+
| encoder | QNN_DLC | float | Qualcomm® SA8775P | 153.502 ms | 1 - 687 MB | NPU
|
| 120 |
+
| encoder | QNN_DLC | float | Qualcomm® SA8775P | 153.502 ms | 1 - 687 MB | NPU
|
| 121 |
+
| encoder | QNN_DLC | float | Qualcomm® SA8775P | 153.502 ms | 1 - 687 MB | NPU
|
| 122 |
+
| encoder | QNN_DLC | float | Qualcomm® QCS9075 | 170.214 ms | 1 - 39 MB | NPU
|
| 123 |
+
| encoder | QNN_DLC | float | Qualcomm® QCS8450 (Proxy) | 269.206 ms | 0 - 823 MB | NPU
|
| 124 |
+
| encoder | QNN_DLC | float | Qualcomm® SA7255P | 437.836 ms | 1 - 696 MB | NPU
|
| 125 |
+
| encoder | QNN_DLC | float | Qualcomm® SA8295P | 192.976 ms | 1 - 611 MB | NPU
|
| 126 |
+
| encoder | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 71.602 ms | 1 - 692 MB | NPU
|
| 127 |
+
| encoder | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 403.997 ms | 42 - 85 MB | GPU
|
| 128 |
+
| encoder | TFLITE | float | Snapdragon® 8 Elite Mobile | 409.903 ms | 40 - 80 MB | GPU
|
| 129 |
+
| encoder | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 475.908 ms | 42 - 184 MB | GPU
|
| 130 |
+
| encoder | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 3135.306 ms | 24 - 69 MB | GPU
|
| 131 |
+
| encoder | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 657.568 ms | 0 - 318 MB | GPU
|
| 132 |
+
| encoder | TFLITE | float | Qualcomm® SA8775P | 1316.653 ms | 20 - 64 MB | GPU
|
| 133 |
+
| encoder | TFLITE | float | Qualcomm® SA8775P | 1316.653 ms | 20 - 64 MB | GPU
|
| 134 |
+
| encoder | TFLITE | float | Qualcomm® SA8775P | 1316.653 ms | 20 - 64 MB | GPU
|
| 135 |
+
| encoder | TFLITE | float | Qualcomm® QCS9075 | 1271.896 ms | 0 - 40 MB | GPU
|
| 136 |
+
| encoder | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 852.126 ms | 39 - 193 MB | GPU
|
| 137 |
+
| encoder | TFLITE | float | Qualcomm® SA7255P | 3135.306 ms | 24 - 69 MB | GPU
|
| 138 |
+
| encoder | TFLITE | float | Qualcomm® SA8295P | 671.062 ms | 38 - 81 MB | GPU
|
| 139 |
+
| encoder | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 409.903 ms | 40 - 80 MB | GPU
|
| 140 |
|
| 141 |
## License
|
| 142 |
* The license for the original implementation of Distil-Whisper can be found
|
release_assets.json
CHANGED
|
@@ -1,26 +1,26 @@
|
|
| 1 |
{
|
| 2 |
-
"version": "0.
|
| 3 |
"precisions": {
|
| 4 |
"float": {
|
| 5 |
"universal_assets": {
|
| 6 |
"tflite": {
|
| 7 |
"tool_versions": {
|
| 8 |
-
"litert": "1.4.
|
| 9 |
},
|
| 10 |
-
"download_url": "https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/distil_whisper/releases/v0.
|
| 11 |
},
|
| 12 |
"qnn_dlc": {
|
| 13 |
"tool_versions": {
|
| 14 |
"qairt": "2.45.0.260326154327"
|
| 15 |
},
|
| 16 |
-
"download_url": "https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/distil_whisper/releases/v0.
|
| 17 |
},
|
| 18 |
"onnx": {
|
| 19 |
"tool_versions": {
|
| 20 |
"qairt": "2.42.0.251225135753_193295",
|
| 21 |
"onnx_runtime": "1.24.3"
|
| 22 |
},
|
| 23 |
-
"download_url": "https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/distil_whisper/releases/v0.
|
| 24 |
}
|
| 25 |
}
|
| 26 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"version": "0.53.0",
|
| 3 |
"precisions": {
|
| 4 |
"float": {
|
| 5 |
"universal_assets": {
|
| 6 |
"tflite": {
|
| 7 |
"tool_versions": {
|
| 8 |
+
"litert": "1.4.3"
|
| 9 |
},
|
| 10 |
+
"download_url": "https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/distil_whisper/releases/v0.53.0/distil_whisper-tflite-float.zip"
|
| 11 |
},
|
| 12 |
"qnn_dlc": {
|
| 13 |
"tool_versions": {
|
| 14 |
"qairt": "2.45.0.260326154327"
|
| 15 |
},
|
| 16 |
+
"download_url": "https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/distil_whisper/releases/v0.53.0/distil_whisper-qnn_dlc-float.zip"
|
| 17 |
},
|
| 18 |
"onnx": {
|
| 19 |
"tool_versions": {
|
| 20 |
"qairt": "2.42.0.251225135753_193295",
|
| 21 |
"onnx_runtime": "1.24.3"
|
| 22 |
},
|
| 23 |
+
"download_url": "https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/distil_whisper/releases/v0.53.0/distil_whisper-onnx-float.zip"
|
| 24 |
}
|
| 25 |
}
|
| 26 |
}
|