Hugging Face's logo

qualcomm
/
OpenAI-Clip

Image Classification
PyTorch
foundation
android
Model card Files
xet
 Community
1
qaihm-bot commited on
Commit
f6c26f4
·
verified ·
1 Parent(s): c39539d

v0.58.0

Browse files

See https://github.com/qualcomm/ai-hub-models/releases/v0.58.0 for changelog.

Files changed (2) hide show
  1. README.md +48 -45
  2. release_assets.json +4 -4
README.md CHANGED
@@ -15,7 +15,7 @@ pipeline_tag: image-classification
15
  Contrastive Language-Image Pre-Training (CLIP) uses a ViT like transformer to get visual features and a causal language model to get the text features. Both the text and visual features can then be used for a variety of zero-shot learning tasks.
16
 
17
  This is based on the implementation of OpenAI-Clip found [here](https://github.com/openai/CLIP/).
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/v0.57.3/src/qai_hub_models/models/openai_clip) 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
 
@@ -28,23 +28,23 @@ Below are pre-exported model assets ready for deployment.
28
 
29
  | Runtime | Precision | Chipset | SDK Versions | Download |
30
  |---|---|---|---|---|
31
- | ONNX | float | Universal | QAIRT 2.45, ONNX Runtime 1.25.0 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/openai_clip/releases/v0.57.3/openai_clip-onnx-float.zip)
32
- | QNN_DLC | float | Universal | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/openai_clip/releases/v0.57.3/openai_clip-qnn_dlc-float.zip)
33
- | TFLITE | float | Universal | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/openai_clip/releases/v0.57.3/openai_clip-tflite-float.zip)
34
 
35
  For more device-specific assets and performance metrics, visit **[OpenAI-Clip on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/openai_clip)**.
36
 
37
 
38
  ### Option 2: Export with Custom Configurations
39
 
40
- Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/v0.57.3/src/qai_hub_models/models/openai_clip) Python library to compile and export the model with your own:
41
  - Custom weights (e.g., fine-tuned checkpoints)
42
  - Custom input shapes
43
  - Target device and runtime configurations
44
 
45
  This option is ideal if you need to customize the model beyond the default configuration provided here.
46
 
47
- See our repository for [OpenAI-Clip on GitHub](https://github.com/qualcomm/ai-hub-models/blob/v0.57.3/src/qai_hub_models/models/openai_clip) for usage instructions.
48
 
49
  ## Model Details
50
 
@@ -60,45 +60,48 @@ See our repository for [OpenAI-Clip on GitHub](https://github.com/qualcomm/ai-hu
60
  ## Performance Summary
61
  | Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
62
  |---|---|---|---|---|---|---
63
- | OpenAI-Clip | ONNX | float | Snapdragon® X2 Elite | 10.356 ms | 180 - 180 MB | NPU
64
- | OpenAI-Clip | ONNX | float | Snapdragon® X Elite | 22.143 ms | 294 - 294 MB | NPU
65
- | OpenAI-Clip | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 14.961 ms | 1 - 1096 MB | NPU
66
- | OpenAI-Clip | ONNX | float | Snapdragon® 8 Gen 1 Mobile | 26.596 ms | 1 - 674 MB | NPU
67
- | OpenAI-Clip | ONNX | float | Qualcomm® QCS8550 (Proxy) | 21.472 ms | 0 - 323 MB | NPU
68
- | OpenAI-Clip | ONNX | float | Qualcomm® QCS8450 | 26.596 ms | 1 - 674 MB | NPU
69
- | OpenAI-Clip | ONNX | float | Snapdragon® 8 Elite Mobile | 11.719 ms | 1 - 870 MB | NPU
70
- | OpenAI-Clip | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 9.258 ms | 1 - 578 MB | NPU
71
- | OpenAI-Clip | ONNX | float | Qualcomm® QCS9075 | 26.356 ms | 1 - 46 MB | NPU
72
- | OpenAI-Clip | ONNX | float | Qualcomm® QCS8750 | 11.719 ms | 1 - 870 MB | NPU
73
- | OpenAI-Clip | ONNX | float | Qualcomm® QCS7181 | 22.143 ms | 294 - 294 MB | NPU
74
- | OpenAI-Clip | QNN_DLC | float | Snapdragon® X2 Elite | 9.648 ms | 1 - 1 MB | NPU
75
- | OpenAI-Clip | QNN_DLC | float | Snapdragon® X Elite | 21.459 ms | 1 - 1 MB | NPU
76
- | OpenAI-Clip | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 14.445 ms | 0 - 768 MB | NPU
77
- | OpenAI-Clip | QNN_DLC | float | Snapdragon® 8 Gen 1 Mobile | 24.142 ms | 0 - 609 MB | NPU
78
- | OpenAI-Clip | QNN_DLC | float | Qualcomm® QCS8275 | 59.355 ms | 1 - 570 MB | NPU
79
- | OpenAI-Clip | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 20.186 ms | 1 - 3 MB | NPU
80
- | OpenAI-Clip | QNN_DLC | float | Qualcomm® QCS8450 | 24.142 ms | 0 - 609 MB | NPU
81
- | OpenAI-Clip | QNN_DLC | float | Snapdragon® 8 Elite Mobile | 11.311 ms | 1 - 587 MB | NPU
82
- | OpenAI-Clip | QNN_DLC | float | Qualcomm® SA8295P | 24.546 ms | 1 - 514 MB | NPU
83
- | OpenAI-Clip | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 8.749 ms | 0 - 419 MB | NPU
84
- | OpenAI-Clip | QNN_DLC | float | Qualcomm® SA7255P | 59.355 ms | 1 - 570 MB | NPU
85
- | OpenAI-Clip | QNN_DLC | float | Qualcomm® QCS9075 | 23.86 ms | 3 - 5 MB | NPU
86
- | OpenAI-Clip | QNN_DLC | float | Qualcomm® QCS8750 | 11.311 ms | 1 - 587 MB | NPU
87
- | OpenAI-Clip | QNN_DLC | float | Qualcomm® QCS7181 | 21.459 ms | 1 - 1 MB | NPU
88
- | OpenAI-Clip | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 14.25 ms | 0 - 775 MB | NPU
89
- | OpenAI-Clip | TFLITE | float | Snapdragon® 8 Gen 1 Mobile | 24.342 ms | 0 - 617 MB | NPU
90
- | OpenAI-Clip | TFLITE | float | Qualcomm® QCS8275 | 59.423 ms | 0 - 566 MB | NPU
91
- | OpenAI-Clip | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 20.105 ms | 0 - 3 MB | NPU
92
- | OpenAI-Clip | TFLITE | float | Qualcomm® SA8775P | 333.229 ms | 6 - 30 MB | CPU
93
- | OpenAI-Clip | TFLITE | float | Qualcomm® SA8650P | 333.229 ms | 6 - 30 MB | CPU
94
- | OpenAI-Clip | TFLITE | float | Qualcomm® SA8255P | 333.229 ms | 6 - 30 MB | CPU
95
- | OpenAI-Clip | TFLITE | float | Qualcomm® QCS8450 | 24.342 ms | 0 - 617 MB | NPU
96
- | OpenAI-Clip | TFLITE | float | Snapdragon® 8 Elite Mobile | 11.342 ms | 0 - 606 MB | NPU
97
- | OpenAI-Clip | TFLITE | float | Qualcomm® SA8295P | 24.559 ms | 0 - 518 MB | NPU
98
- | OpenAI-Clip | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 8.74 ms | 0 - 433 MB | NPU
99
- | OpenAI-Clip | TFLITE | float | Qualcomm® SA7255P | 59.423 ms | 0 - 566 MB | NPU
100
- | OpenAI-Clip | TFLITE | float | Qualcomm® QCS9075 | 23.106 ms | 0 - 296 MB | NPU
101
- | OpenAI-Clip | TFLITE | float | Qualcomm® QCS8750 | 11.342 ms | 0 - 606 MB | NPU
 
 
 
102
 
103
  ## License
104
  * The license for the original implementation of OpenAI-Clip can be found
 
15
  Contrastive Language-Image Pre-Training (CLIP) uses a ViT like transformer to get visual features and a causal language model to get the text features. Both the text and visual features can then be used for a variety of zero-shot learning tasks.
16
 
17
  This is based on the implementation of OpenAI-Clip found [here](https://github.com/openai/CLIP/).
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/v0.58.0/src/qai_hub_models/models/openai_clip) 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 2.45, ONNX Runtime 1.25.0 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/openai_clip/releases/v0.58.0/openai_clip-onnx-float.zip)
32
+ | QNN_DLC | float | Universal | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/openai_clip/releases/v0.58.0/openai_clip-qnn_dlc-float.zip)
33
+ | TFLITE | float | Universal | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/openai_clip/releases/v0.58.0/openai_clip-tflite-float.zip)
34
 
35
  For more device-specific assets and performance metrics, visit **[OpenAI-Clip on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/openai_clip)**.
36
 
37
 
38
  ### Option 2: Export with Custom Configurations
39
 
40
+ Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/v0.58.0/src/qai_hub_models/models/openai_clip) Python library to compile and export the model with your own:
41
  - Custom weights (e.g., fine-tuned checkpoints)
42
  - Custom input shapes
43
  - Target device and runtime configurations
44
 
45
  This option is ideal if you need to customize the model beyond the default configuration provided here.
46
 
47
+ See our repository for [OpenAI-Clip on GitHub](https://github.com/qualcomm/ai-hub-models/blob/v0.58.0/src/qai_hub_models/models/openai_clip) for usage instructions.
48
 
49
  ## Model Details
50
 
 
60
  ## Performance Summary
61
  | Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
62
  |---|---|---|---|---|---|---
63
+ | OpenAI-Clip | ONNX | float | Snapdragon® X2 Elite | 10.389 ms | 2 - 2 MB | NPU
64
+ | OpenAI-Clip | ONNX | float | Snapdragon® X Elite | 22.087 ms | 295 - 295 MB | NPU
65
+ | OpenAI-Clip | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 14.92 ms | 0 - 1105 MB | NPU
66
+ | OpenAI-Clip | ONNX | float | Snapdragon® 8 Gen 1 Mobile | 26.753 ms | 1 - 687 MB | NPU
67
+ | OpenAI-Clip | ONNX | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 21.325 ms | 0 - 323 MB | NPU
68
+ | OpenAI-Clip | ONNX | float | Qualcomm® QCS8450 | 26.753 ms | 1 - 687 MB | NPU
69
+ | OpenAI-Clip | ONNX | float | Qualcomm® Dragonwing™ IQ-9075 | 30.849 ms | 0 - 4 MB | NPU
70
+ | OpenAI-Clip | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 9.261 ms | 1 - 584 MB | NPU
71
+ | OpenAI-Clip | ONNX | float | Snapdragon® 8 Elite Mobile | 11.631 ms | 0 - 869 MB | NPU
72
+ | OpenAI-Clip | ONNX | float | Qualcomm® Dragonwing™ Q-8750 | 11.631 ms | 0 - 869 MB | NPU
73
+ | OpenAI-Clip | ONNX | float | Qualcomm® Dragonwing™ IQ-X7181 | 22.087 ms | 295 - 295 MB | NPU
74
+ | OpenAI-Clip | QNN_DLC | float | Snapdragon® X2 Elite | 9.645 ms | 1 - 1 MB | NPU
75
+ | OpenAI-Clip | QNN_DLC | float | Snapdragon® X Elite | 21.391 ms | 1 - 1 MB | NPU
76
+ | OpenAI-Clip | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 14.547 ms | 0 - 770 MB | NPU
77
+ | OpenAI-Clip | QNN_DLC | float | Snapdragon® 8 Gen 1 Mobile | 24.245 ms | 0 - 608 MB | NPU
78
+ | OpenAI-Clip | QNN_DLC | float | Qualcomm® QCS8275 | 59.346 ms | 1 - 569 MB | NPU
79
+ | OpenAI-Clip | QNN_DLC | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 20.411 ms | 1 - 579 MB | NPU
80
+ | OpenAI-Clip | QNN_DLC | float | Qualcomm® SA8775P | 23.09 ms | 1 - 568 MB | NPU
81
+ | OpenAI-Clip | QNN_DLC | float | Qualcomm® SA8650P | 23.09 ms | 1 - 568 MB | NPU
82
+ | OpenAI-Clip | QNN_DLC | float | Qualcomm® SA8255P | 23.09 ms | 1 - 568 MB | NPU
83
+ | OpenAI-Clip | QNN_DLC | float | Qualcomm® QCS8450 | 24.245 ms | 0 - 608 MB | NPU
84
+ | OpenAI-Clip | QNN_DLC | float | Qualcomm® Dragonwing™ IQ-9075 | 23.86 ms | 3 - 5 MB | NPU
85
+ | OpenAI-Clip | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 8.742 ms | 0 - 419 MB | NPU
86
+ | OpenAI-Clip | QNN_DLC | float | Qualcomm® SA7255P | 59.346 ms | 1 - 569 MB | NPU
87
+ | OpenAI-Clip | QNN_DLC | float | Snapdragon® 8 Elite Mobile | 11.321 ms | 0 - 586 MB | NPU
88
+ | OpenAI-Clip | QNN_DLC | float | Qualcomm® SA8295P | 24.62 ms | 1 - 514 MB | NPU
89
+ | OpenAI-Clip | QNN_DLC | float | Qualcomm® Dragonwing™ Q-8750 | 11.321 ms | 0 - 586 MB | NPU
90
+ | OpenAI-Clip | QNN_DLC | float | Qualcomm® Dragonwing™ IQ-X7181 | 21.391 ms | 1 - 1 MB | NPU
91
+ | OpenAI-Clip | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 14.216 ms | 0 - 773 MB | NPU
92
+ | OpenAI-Clip | TFLITE | float | Snapdragon® 8 Gen 1 Mobile | 24.231 ms | 0 - 618 MB | NPU
93
+ | OpenAI-Clip | TFLITE | float | Qualcomm® QCS8275 | 59.398 ms | 0 - 565 MB | NPU
94
+ | OpenAI-Clip | TFLITE | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 20.306 ms | 0 - 3 MB | NPU
95
+ | OpenAI-Clip | TFLITE | float | Qualcomm® SA8775P | 23.093 ms | 0 - 566 MB | NPU
96
+ | OpenAI-Clip | TFLITE | float | Qualcomm® SA8650P | 23.093 ms | 0 - 566 MB | NPU
97
+ | OpenAI-Clip | TFLITE | float | Qualcomm® SA8255P | 23.093 ms | 0 - 566 MB | NPU
98
+ | OpenAI-Clip | TFLITE | float | Qualcomm® QCS8450 | 24.231 ms | 0 - 618 MB | NPU
99
+ | OpenAI-Clip | TFLITE | float | Qualcomm® Dragonwing™ IQ-9075 | 23.106 ms | 0 - 296 MB | NPU
100
+ | OpenAI-Clip | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 8.745 ms | 0 - 434 MB | NPU
101
+ | OpenAI-Clip | TFLITE | float | Qualcomm® SA7255P | 59.398 ms | 0 - 565 MB | NPU
102
+ | OpenAI-Clip | TFLITE | float | Snapdragon® 8 Elite Mobile | 11.303 ms | 0 - 599 MB | NPU
103
+ | OpenAI-Clip | TFLITE | float | Qualcomm® SA8295P | 24.52 ms | 0 - 517 MB | NPU
104
+ | OpenAI-Clip | TFLITE | float | Qualcomm® Dragonwing™ Q-8750 | 11.303 ms | 0 - 599 MB | NPU
105
 
106
  ## License
107
  * The license for the original implementation of OpenAI-Clip can be found
release_assets.json CHANGED
@@ -1,5 +1,5 @@
1
  {
2
- "version": "0.57.3",
3
  "precisions": {
4
  "float": {
5
  "universal_assets": {
@@ -8,20 +8,20 @@
8
  "qairt": "2.45.0.260326154327",
9
  "litert": "1.4.4"
10
  },
11
- "download_url": "https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/openai_clip/releases/v0.57.3/openai_clip-tflite-float.zip"
12
  },
13
  "qnn_dlc": {
14
  "tool_versions": {
15
  "qairt": "2.45.0.260326154327"
16
  },
17
- "download_url": "https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/openai_clip/releases/v0.57.3/openai_clip-qnn_dlc-float.zip"
18
  },
19
  "onnx": {
20
  "tool_versions": {
21
  "qairt": "2.45.0.260326154327",
22
  "onnx_runtime": "1.25.0"
23
  },
24
- "download_url": "https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/openai_clip/releases/v0.57.3/openai_clip-onnx-float.zip"
25
  }
26
  }
27
  }
 
1
  {
2
+ "version": "0.58.0",
3
  "precisions": {
4
  "float": {
5
  "universal_assets": {
 
8
  "qairt": "2.45.0.260326154327",
9
  "litert": "1.4.4"
10
  },
11
+ "download_url": "https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/openai_clip/releases/v0.58.0/openai_clip-tflite-float.zip"
12
  },
13
  "qnn_dlc": {
14
  "tool_versions": {
15
  "qairt": "2.45.0.260326154327"
16
  },
17
+ "download_url": "https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/openai_clip/releases/v0.58.0/openai_clip-qnn_dlc-float.zip"
18
  },
19
  "onnx": {
20
  "tool_versions": {
21
  "qairt": "2.45.0.260326154327",
22
  "onnx_runtime": "1.25.0"
23
  },
24
+ "download_url": "https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/openai_clip/releases/v0.58.0/openai_clip-onnx-float.zip"
25
  }
26
  }
27
  }