Image Classification
PyTorch
bu_auto
android
qaihm-bot commited on
Commit
48e2967
·
verified ·
1 Parent(s): 9724414

See https://github.com/qualcomm/ai-hub-models/releases/v0.65.0 for changelog.

Files changed (2) hide show
  1. README.md +9 -36
  2. release_assets.json +7 -24
README.md CHANGED
@@ -15,7 +15,7 @@ pipeline_tag: image-classification
15
  LeViT is a vision transformer model that can classify images from the Imagenet dataset.
16
 
17
  This is based on the implementation of LeViT found [here](https://github.com/facebookresearch/LeViT).
18
- This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/v0.64.0/src/qai_hub_models/models/levit) library to export with custom configurations. More details on model performance across various devices, can be found [here](#performance-summary).
19
 
20
  Qualcomm AI Hub Models uses [Qualcomm AI Hub Workbench](https://workbench.aihub.qualcomm.com) to compile, profile, and evaluate this model. [Sign up](https://myaccount.qualcomm.com/signup) to run these models on a hosted Qualcomm® device.
21
 
@@ -28,28 +28,26 @@ Below are pre-exported model assets ready for deployment.
28
 
29
  | Runtime | Precision | Chipset | SDK Versions | Download |
30
  |---|---|---|---|---|
31
- | ONNX | float | Universal | QAIRT 2.50, ONNX Runtime 1.30.0 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/levit/releases/v0.64.0/levit-onnx-float.zip)
32
- | ONNX | w8a16 | Universal | QAIRT 2.50, ONNX Runtime 1.30.0 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/levit/releases/v0.64.0/levit-onnx-w8a16.zip)
33
- | ONNX | w8a16_mixed_int16 | Universal | QAIRT 2.45, ONNX Runtime 1.27.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/levit/releases/v0.64.0/levit-onnx-w8a16_mixed_int16.zip)
34
- | ONNX | w8a8_mixed_int16 | Universal | QAIRT 2.50, ONNX Runtime 1.30.0 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/levit/releases/v0.64.0/levit-onnx-w8a8_mixed_int16.zip)
35
- | QNN_DLC | float | Universal | QAIRT 2.50 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/levit/releases/v0.64.0/levit-qnn_dlc-float.zip)
36
- | QNN_DLC | w8a16 | Universal | QAIRT 2.50 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/levit/releases/v0.64.0/levit-qnn_dlc-w8a16.zip)
37
- | QNN_DLC | w8a16_mixed_int16 | Universal | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/levit/releases/v0.64.0/levit-qnn_dlc-w8a16_mixed_int16.zip)
38
- | TFLITE | float | Universal | QAIRT 2.50 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/levit/releases/v0.64.0/levit-tflite-float.zip)
39
 
40
  For more device-specific assets and performance metrics, visit **[LeViT on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/levit)**.
41
 
42
 
43
  ### Option 2: Export with Custom Configurations
44
 
45
- Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/v0.64.0/src/qai_hub_models/models/levit) Python library to compile and export the model with your own:
46
  - Custom weights (e.g., fine-tuned checkpoints)
47
  - Custom input shapes
48
  - Target device and runtime configurations
49
 
50
  This option is ideal if you need to customize the model beyond the default configuration provided here.
51
 
52
- See our repository for [LeViT on GitHub](https://github.com/qualcomm/ai-hub-models/blob/v0.64.0/src/qai_hub_models/models/levit) for usage instructions.
53
 
54
  ## Model Details
55
 
@@ -93,15 +91,6 @@ See our repository for [LeViT on GitHub](https://github.com/qualcomm/ai-hub-mode
93
  | LeViT | ONNX | w8a16 | Qualcomm® Dragonwing™ Q-7790 | 0.995 ms | 0 - 60 MB | NPU
94
  | LeViT | ONNX | w8a16 | Qualcomm® Dragonwing™ Q-8750 | 0.513 ms | 0 - 63 MB | NPU
95
  | LeViT | ONNX | w8a16 | Snapdragon® 7 Gen 4 Mobile | 0.995 ms | 0 - 60 MB | NPU
96
- | LeViT | ONNX | w8a16_mixed_int16 | Snapdragon® 8 Elite Gen 5 For Galaxy Mobile | 0.497 ms | 0 - 63 MB | NPU
97
- | LeViT | ONNX | w8a16_mixed_int16 | Snapdragon® 8 Elite For Galaxy Mobile | 0.549 ms | 0 - 55 MB | NPU
98
- | LeViT | ONNX | w8a16_mixed_int16 | Snapdragon® X2 Elite | 0.509 ms | 1 - 1 MB | NPU
99
- | LeViT | ONNX | w8a16_mixed_int16 | Snapdragon® X Elite | 1.114 ms | 10 - 10 MB | NPU
100
- | LeViT | ONNX | w8a16_mixed_int16 | Snapdragon® 8 Gen 3 Mobile | 0.721 ms | 0 - 85 MB | NPU
101
- | LeViT | ONNX | w8a16_mixed_int16 | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 1.069 ms | 0 - 58 MB | NPU
102
- | LeViT | ONNX | w8a16_mixed_int16 | Qualcomm® Dragonwing™ IQ-9075 | 1.217 ms | 0 - 3 MB | NPU
103
- | LeViT | ONNX | w8a16_mixed_int16 | Qualcomm® Dragonwing™ IQ-X7181 | 1.114 ms | 10 - 10 MB | NPU
104
- | LeViT | ONNX | w8a16_mixed_int16 | Qualcomm® Dragonwing™ Q-8750 | 0.549 ms | 0 - 55 MB | NPU
105
  | LeViT | ONNX | w8a8_mixed_int16 | Snapdragon® 8 Elite Gen 5 For Galaxy Mobile | 1.476 ms | 0 - 58 MB | NPU
106
  | LeViT | ONNX | w8a8_mixed_int16 | Snapdragon® 8 Elite For Galaxy Mobile | 2.118 ms | 0 - 58 MB | NPU
107
  | LeViT | ONNX | w8a8_mixed_int16 | Snapdragon® 8 Gen 3 Mobile | 1.953 ms | 0 - 85 MB | NPU
@@ -139,22 +128,6 @@ See our repository for [LeViT on GitHub](https://github.com/qualcomm/ai-hub-mode
139
  | LeViT | QNN_DLC | w8a16 | Qualcomm® Dragonwing™ Q-8750 | 0.681 ms | 0 - 52 MB | NPU
140
  | LeViT | QNN_DLC | w8a16 | Qualcomm® SA7255P | 2.737 ms | 0 - 55 MB | NPU
141
  | LeViT | QNN_DLC | w8a16 | Snapdragon® 7 Gen 4 Mobile | 1.402 ms | 0 - 56 MB | NPU
142
- | LeViT | QNN_DLC | w8a16_mixed_int16 | Snapdragon® 8 Elite Gen 5 For Galaxy Mobile | 0.604 ms | 0 - 57 MB | NPU
143
- | LeViT | QNN_DLC | w8a16_mixed_int16 | Snapdragon® 8 Elite For Galaxy Mobile | 0.698 ms | 0 - 49 MB | NPU
144
- | LeViT | QNN_DLC | w8a16_mixed_int16 | Snapdragon® X2 Elite | 0.801 ms | 0 - 0 MB | NPU
145
- | LeViT | QNN_DLC | w8a16_mixed_int16 | Snapdragon® X Elite | 1.652 ms | 0 - 0 MB | NPU
146
- | LeViT | QNN_DLC | w8a16_mixed_int16 | Snapdragon® 8 Gen 3 Mobile | 0.973 ms | 0 - 75 MB | NPU
147
- | LeViT | QNN_DLC | w8a16_mixed_int16 | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 1.451 ms | 0 - 3 MB | NPU
148
- | LeViT | QNN_DLC | w8a16_mixed_int16 | Qualcomm® SA8775P | 1.759 ms | 0 - 53 MB | NPU
149
- | LeViT | QNN_DLC | w8a16_mixed_int16 | Qualcomm® SA8650P | 1.759 ms | 0 - 53 MB | NPU
150
- | LeViT | QNN_DLC | w8a16_mixed_int16 | Qualcomm® SA8255P | 1.759 ms | 0 - 53 MB | NPU
151
- | LeViT | QNN_DLC | w8a16_mixed_int16 | Qualcomm® Dragonwing™ IQ-9075 | 1.665 ms | 0 - 2 MB | NPU
152
- | LeViT | QNN_DLC | w8a16_mixed_int16 | Qualcomm® Dragonwing™ IQ-X7181 | 1.652 ms | 0 - 0 MB | NPU
153
- | LeViT | QNN_DLC | w8a16_mixed_int16 | Qualcomm® Dragonwing™ Q-6690 | 5.938 ms | 0 - 177 MB | NPU
154
- | LeViT | QNN_DLC | w8a16_mixed_int16 | Qualcomm® Dragonwing™ Q-7790 | 1.435 ms | 0 - 51 MB | NPU
155
- | LeViT | QNN_DLC | w8a16_mixed_int16 | Qualcomm® Dragonwing™ Q-8750 | 0.698 ms | 0 - 49 MB | NPU
156
- | LeViT | QNN_DLC | w8a16_mixed_int16 | Qualcomm® SA7255P | 2.877 ms | 0 - 51 MB | NPU
157
- | LeViT | QNN_DLC | w8a16_mixed_int16 | Snapdragon® 7 Gen 4 Mobile | 1.435 ms | 0 - 51 MB | NPU
158
  | LeViT | TFLITE | float | Snapdragon® 8 Elite Gen 5 For Galaxy Mobile | 0.612 ms | 0 - 58 MB | NPU
159
  | LeViT | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.722 ms | 0 - 61 MB | NPU
160
  | LeViT | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 0.953 ms | 0 - 85 MB | NPU
 
15
  LeViT is a vision transformer model that can classify images from the Imagenet dataset.
16
 
17
  This is based on the implementation of LeViT found [here](https://github.com/facebookresearch/LeViT).
18
+ This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/v0.65.0/src/qai_hub_models/models/levit) library to export with custom configurations. More details on model performance across various devices, can be found [here](#performance-summary).
19
 
20
  Qualcomm AI Hub Models uses [Qualcomm AI Hub Workbench](https://workbench.aihub.qualcomm.com) to compile, profile, and evaluate this model. [Sign up](https://myaccount.qualcomm.com/signup) to run these models on a hosted Qualcomm® device.
21
 
 
28
 
29
  | Runtime | Precision | Chipset | SDK Versions | Download |
30
  |---|---|---|---|---|
31
+ | ONNX | float | Universal | QAIRT 2.50, ONNX Runtime 1.30.0 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/levit/releases/v0.65.0/levit-onnx-float.zip)
32
+ | ONNX | w8a16 | Universal | QAIRT 2.50, ONNX Runtime 1.30.0 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/levit/releases/v0.65.0/levit-onnx-w8a16.zip)
33
+ | ONNX | w8a8_mixed_int16 | Universal | QAIRT 2.50, ONNX Runtime 1.30.0 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/levit/releases/v0.65.0/levit-onnx-w8a8_mixed_int16.zip)
34
+ | QNN_DLC | float | Universal | QAIRT 2.50 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/levit/releases/v0.65.0/levit-qnn_dlc-float.zip)
35
+ | QNN_DLC | w8a16 | Universal | QAIRT 2.50 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/levit/releases/v0.65.0/levit-qnn_dlc-w8a16.zip)
36
+ | TFLITE | float | Universal | QAIRT 2.50 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/levit/releases/v0.65.0/levit-tflite-float.zip)
 
 
37
 
38
  For more device-specific assets and performance metrics, visit **[LeViT on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/levit)**.
39
 
40
 
41
  ### Option 2: Export with Custom Configurations
42
 
43
+ Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/v0.65.0/src/qai_hub_models/models/levit) Python library to compile and export the model with your own:
44
  - Custom weights (e.g., fine-tuned checkpoints)
45
  - Custom input shapes
46
  - Target device and runtime configurations
47
 
48
  This option is ideal if you need to customize the model beyond the default configuration provided here.
49
 
50
+ See our repository for [LeViT on GitHub](https://github.com/qualcomm/ai-hub-models/blob/v0.65.0/src/qai_hub_models/models/levit) for usage instructions.
51
 
52
  ## Model Details
53
 
 
91
  | LeViT | ONNX | w8a16 | Qualcomm® Dragonwing™ Q-7790 | 0.995 ms | 0 - 60 MB | NPU
92
  | LeViT | ONNX | w8a16 | Qualcomm® Dragonwing™ Q-8750 | 0.513 ms | 0 - 63 MB | NPU
93
  | LeViT | ONNX | w8a16 | Snapdragon® 7 Gen 4 Mobile | 0.995 ms | 0 - 60 MB | NPU
 
 
 
 
 
 
 
 
 
94
  | LeViT | ONNX | w8a8_mixed_int16 | Snapdragon® 8 Elite Gen 5 For Galaxy Mobile | 1.476 ms | 0 - 58 MB | NPU
95
  | LeViT | ONNX | w8a8_mixed_int16 | Snapdragon® 8 Elite For Galaxy Mobile | 2.118 ms | 0 - 58 MB | NPU
96
  | LeViT | ONNX | w8a8_mixed_int16 | Snapdragon® 8 Gen 3 Mobile | 1.953 ms | 0 - 85 MB | NPU
 
128
  | LeViT | QNN_DLC | w8a16 | Qualcomm® Dragonwing™ Q-8750 | 0.681 ms | 0 - 52 MB | NPU
129
  | LeViT | QNN_DLC | w8a16 | Qualcomm® SA7255P | 2.737 ms | 0 - 55 MB | NPU
130
  | LeViT | QNN_DLC | w8a16 | Snapdragon® 7 Gen 4 Mobile | 1.402 ms | 0 - 56 MB | NPU
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
131
  | LeViT | TFLITE | float | Snapdragon® 8 Elite Gen 5 For Galaxy Mobile | 0.612 ms | 0 - 58 MB | NPU
132
  | LeViT | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.722 ms | 0 - 61 MB | NPU
133
  | LeViT | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 0.953 ms | 0 - 85 MB | NPU
release_assets.json CHANGED
@@ -1,5 +1,5 @@
1
  {
2
- "version": "0.64.0",
3
  "precisions": {
4
  "float": {
5
  "universal_assets": {
@@ -8,19 +8,19 @@
8
  "qairt": "2.50.0.260828221209",
9
  "onnx_runtime": "1.30.0"
10
  },
11
- "download_url": "https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/levit/releases/v0.64.0/levit-onnx-float.zip"
12
  },
13
  "qnn_dlc": {
14
  "tool_versions": {
15
  "qairt": "2.50.0.260828221209"
16
  },
17
- "download_url": "https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/levit/releases/v0.64.0/levit-qnn_dlc-float.zip"
18
  },
19
  "tflite": {
20
  "tool_versions": {
21
  "qairt": "2.50.0.260828221209"
22
  },
23
- "download_url": "https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/levit/releases/v0.64.0/levit-tflite-float.zip"
24
  }
25
  }
26
  },
@@ -31,30 +31,13 @@
31
  "qairt": "2.50.0.260828221209",
32
  "onnx_runtime": "1.30.0"
33
  },
34
- "download_url": "https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/levit/releases/v0.64.0/levit-onnx-w8a16.zip"
35
  },
36
  "qnn_dlc": {
37
  "tool_versions": {
38
  "qairt": "2.50.0.260828221209"
39
  },
40
- "download_url": "https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/levit/releases/v0.64.0/levit-qnn_dlc-w8a16.zip"
41
- }
42
- }
43
- },
44
- "w8a16_mixed_int16": {
45
- "universal_assets": {
46
- "onnx": {
47
- "tool_versions": {
48
- "qairt": "2.45.0.260326154327",
49
- "onnx_runtime": "1.27.1"
50
- },
51
- "download_url": "https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/levit/releases/v0.64.0/levit-onnx-w8a16_mixed_int16.zip"
52
- },
53
- "qnn_dlc": {
54
- "tool_versions": {
55
- "qairt": "2.45.0.260326154327"
56
- },
57
- "download_url": "https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/levit/releases/v0.64.0/levit-qnn_dlc-w8a16_mixed_int16.zip"
58
  }
59
  }
60
  },
@@ -65,7 +48,7 @@
65
  "qairt": "2.50.0.260828221209",
66
  "onnx_runtime": "1.30.0"
67
  },
68
- "download_url": "https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/levit/releases/v0.64.0/levit-onnx-w8a8_mixed_int16.zip"
69
  }
70
  }
71
  }
 
1
  {
2
+ "version": "0.65.0",
3
  "precisions": {
4
  "float": {
5
  "universal_assets": {
 
8
  "qairt": "2.50.0.260828221209",
9
  "onnx_runtime": "1.30.0"
10
  },
11
+ "download_url": "https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/levit/releases/v0.65.0/levit-onnx-float.zip"
12
  },
13
  "qnn_dlc": {
14
  "tool_versions": {
15
  "qairt": "2.50.0.260828221209"
16
  },
17
+ "download_url": "https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/levit/releases/v0.65.0/levit-qnn_dlc-float.zip"
18
  },
19
  "tflite": {
20
  "tool_versions": {
21
  "qairt": "2.50.0.260828221209"
22
  },
23
+ "download_url": "https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/levit/releases/v0.65.0/levit-tflite-float.zip"
24
  }
25
  }
26
  },
 
31
  "qairt": "2.50.0.260828221209",
32
  "onnx_runtime": "1.30.0"
33
  },
34
+ "download_url": "https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/levit/releases/v0.65.0/levit-onnx-w8a16.zip"
35
  },
36
  "qnn_dlc": {
37
  "tool_versions": {
38
  "qairt": "2.50.0.260828221209"
39
  },
40
+ "download_url": "https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/levit/releases/v0.65.0/levit-qnn_dlc-w8a16.zip"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
41
  }
42
  }
43
  },
 
48
  "qairt": "2.50.0.260828221209",
49
  "onnx_runtime": "1.30.0"
50
  },
51
+ "download_url": "https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/levit/releases/v0.65.0/levit-onnx-w8a8_mixed_int16.zip"
52
  }
53
  }
54
  }