|
--- |
|
library_name: pytorch |
|
license: mit |
|
pipeline_tag: image-classification |
|
tags: |
|
- foundation |
|
- android |
|
|
|
--- |
|
|
|
 |
|
|
|
# OpenAI-Clip: Optimized for Mobile Deployment |
|
## Multi-modal foundational model for vision and language tasks like image/text similarity and for zero-shot image classification |
|
|
|
|
|
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. |
|
|
|
This model is an implementation of OpenAI-Clip found [here](https://github.com/openai/CLIP/). |
|
|
|
|
|
This repository provides scripts to run OpenAI-Clip on Qualcomm® devices. |
|
More details on model performance across various devices, can be found |
|
[here](https://aihub.qualcomm.com/models/openai_clip). |
|
|
|
|
|
### Model Details |
|
|
|
- **Model Type:** Image classification |
|
- **Model Stats:** |
|
- Model checkpoint: ViT-B/16 |
|
- Image input resolution: 224x224 |
|
- Text context length: 77 |
|
- Number of parameters (CLIPTextEncoder): 76.0M |
|
- Model size (CLIPTextEncoder): 290 MB |
|
- Number of parameters (CLIPImageEncoder): 115M |
|
- Model size (CLIPImageEncoder): 437 MB |
|
|
|
| Model | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model |
|
|---|---|---|---|---|---|---|---|---| |
|
| CLIPImageEncoder | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 34.319 ms | 0 - 54 MB | FP16 | NPU | [OpenAI-Clip.tflite](https://huggingface.co./qualcomm/OpenAI-Clip/blob/main/CLIPImageEncoder.tflite) | |
|
| CLIPImageEncoder | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 26.337 ms | 0 - 56 MB | FP16 | NPU | [OpenAI-Clip.so](https://huggingface.co./qualcomm/OpenAI-Clip/blob/main/CLIPImageEncoder.so) | |
|
| CLIPImageEncoder | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 27.11 ms | 0 - 265 MB | FP16 | NPU | [OpenAI-Clip.tflite](https://huggingface.co./qualcomm/OpenAI-Clip/blob/main/CLIPImageEncoder.tflite) | |
|
| CLIPImageEncoder | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 20.295 ms | 0 - 172 MB | FP16 | NPU | [OpenAI-Clip.so](https://huggingface.co./qualcomm/OpenAI-Clip/blob/main/CLIPImageEncoder.so) | |
|
| CLIPImageEncoder | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 25.123 ms | 0 - 265 MB | FP16 | NPU | [OpenAI-Clip.tflite](https://huggingface.co./qualcomm/OpenAI-Clip/blob/main/CLIPImageEncoder.tflite) | |
|
| CLIPImageEncoder | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 16.937 ms | 1 - 173 MB | FP16 | NPU | Use Export Script | |
|
| CLIPImageEncoder | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 34.064 ms | 0 - 63 MB | FP16 | NPU | [OpenAI-Clip.tflite](https://huggingface.co./qualcomm/OpenAI-Clip/blob/main/CLIPImageEncoder.tflite) | |
|
| CLIPImageEncoder | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 20.375 ms | 1 - 4 MB | FP16 | NPU | Use Export Script | |
|
| CLIPImageEncoder | SA7255P ADP | SA7255P | TFLITE | 326.663 ms | 0 - 264 MB | FP16 | NPU | [OpenAI-Clip.tflite](https://huggingface.co./qualcomm/OpenAI-Clip/blob/main/CLIPImageEncoder.tflite) | |
|
| CLIPImageEncoder | SA7255P ADP | SA7255P | QNN | 264.763 ms | 1 - 9 MB | FP16 | NPU | Use Export Script | |
|
| CLIPImageEncoder | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 34.194 ms | 0 - 56 MB | FP16 | NPU | [OpenAI-Clip.tflite](https://huggingface.co./qualcomm/OpenAI-Clip/blob/main/CLIPImageEncoder.tflite) | |
|
| CLIPImageEncoder | SA8255 (Proxy) | SA8255P Proxy | QNN | 20.535 ms | 1 - 3 MB | FP16 | NPU | Use Export Script | |
|
| CLIPImageEncoder | SA8295P ADP | SA8295P | TFLITE | 40.104 ms | 0 - 202 MB | FP16 | NPU | [OpenAI-Clip.tflite](https://huggingface.co./qualcomm/OpenAI-Clip/blob/main/CLIPImageEncoder.tflite) | |
|
| CLIPImageEncoder | SA8295P ADP | SA8295P | QNN | 30.975 ms | 1 - 15 MB | FP16 | NPU | Use Export Script | |
|
| CLIPImageEncoder | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 34.131 ms | 0 - 50 MB | FP16 | NPU | [OpenAI-Clip.tflite](https://huggingface.co./qualcomm/OpenAI-Clip/blob/main/CLIPImageEncoder.tflite) | |
|
| CLIPImageEncoder | SA8650 (Proxy) | SA8650P Proxy | QNN | 20.714 ms | 1 - 3 MB | FP16 | NPU | Use Export Script | |
|
| CLIPImageEncoder | SA8775P ADP | SA8775P | TFLITE | 42.575 ms | 0 - 264 MB | FP16 | NPU | [OpenAI-Clip.tflite](https://huggingface.co./qualcomm/OpenAI-Clip/blob/main/CLIPImageEncoder.tflite) | |
|
| CLIPImageEncoder | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 35.825 ms | 0 - 202 MB | FP16 | NPU | [OpenAI-Clip.tflite](https://huggingface.co./qualcomm/OpenAI-Clip/blob/main/CLIPImageEncoder.tflite) | |
|
| CLIPImageEncoder | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 29.758 ms | 0 - 166 MB | FP16 | NPU | Use Export Script | |
|
| CLIPImageEncoder | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 22.195 ms | 1 - 1 MB | FP16 | NPU | Use Export Script | |
|
| CLIPTextEncoder | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 5.68 ms | 0 - 24 MB | FP16 | NPU | [OpenAI-Clip.tflite](https://huggingface.co./qualcomm/OpenAI-Clip/blob/main/CLIPTextEncoder.tflite) | |
|
| CLIPTextEncoder | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 4.678 ms | 0 - 25 MB | FP16 | NPU | [OpenAI-Clip.so](https://huggingface.co./qualcomm/OpenAI-Clip/blob/main/CLIPTextEncoder.so) | |
|
| CLIPTextEncoder | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 3.993 ms | 0 - 88 MB | FP16 | NPU | [OpenAI-Clip.tflite](https://huggingface.co./qualcomm/OpenAI-Clip/blob/main/CLIPTextEncoder.tflite) | |
|
| CLIPTextEncoder | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 3.297 ms | 0 - 71 MB | FP16 | NPU | [OpenAI-Clip.so](https://huggingface.co./qualcomm/OpenAI-Clip/blob/main/CLIPTextEncoder.so) | |
|
| CLIPTextEncoder | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 3.98 ms | 0 - 85 MB | FP16 | NPU | [OpenAI-Clip.tflite](https://huggingface.co./qualcomm/OpenAI-Clip/blob/main/CLIPTextEncoder.tflite) | |
|
| CLIPTextEncoder | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 3.232 ms | 0 - 70 MB | FP16 | NPU | Use Export Script | |
|
| CLIPTextEncoder | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 5.699 ms | 0 - 18 MB | FP16 | NPU | [OpenAI-Clip.tflite](https://huggingface.co./qualcomm/OpenAI-Clip/blob/main/CLIPTextEncoder.tflite) | |
|
| CLIPTextEncoder | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 4.768 ms | 0 - 3 MB | FP16 | NPU | Use Export Script | |
|
| CLIPTextEncoder | SA7255P ADP | SA7255P | TFLITE | 61.254 ms | 0 - 81 MB | FP16 | NPU | [OpenAI-Clip.tflite](https://huggingface.co./qualcomm/OpenAI-Clip/blob/main/CLIPTextEncoder.tflite) | |
|
| CLIPTextEncoder | SA7255P ADP | SA7255P | QNN | 51.477 ms | 0 - 9 MB | FP16 | NPU | Use Export Script | |
|
| CLIPTextEncoder | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 5.722 ms | 0 - 24 MB | FP16 | NPU | [OpenAI-Clip.tflite](https://huggingface.co./qualcomm/OpenAI-Clip/blob/main/CLIPTextEncoder.tflite) | |
|
| CLIPTextEncoder | SA8255 (Proxy) | SA8255P Proxy | QNN | 4.804 ms | 0 - 2 MB | FP16 | NPU | Use Export Script | |
|
| CLIPTextEncoder | SA8295P ADP | SA8295P | TFLITE | 7.627 ms | 0 - 69 MB | FP16 | NPU | [OpenAI-Clip.tflite](https://huggingface.co./qualcomm/OpenAI-Clip/blob/main/CLIPTextEncoder.tflite) | |
|
| CLIPTextEncoder | SA8295P ADP | SA8295P | QNN | 6.596 ms | 0 - 14 MB | FP16 | NPU | Use Export Script | |
|
| CLIPTextEncoder | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 5.726 ms | 0 - 18 MB | FP16 | NPU | [OpenAI-Clip.tflite](https://huggingface.co./qualcomm/OpenAI-Clip/blob/main/CLIPTextEncoder.tflite) | |
|
| CLIPTextEncoder | SA8650 (Proxy) | SA8650P Proxy | QNN | 4.811 ms | 0 - 3 MB | FP16 | NPU | Use Export Script | |
|
| CLIPTextEncoder | SA8775P ADP | SA8775P | TFLITE | 8.166 ms | 0 - 81 MB | FP16 | NPU | [OpenAI-Clip.tflite](https://huggingface.co./qualcomm/OpenAI-Clip/blob/main/CLIPTextEncoder.tflite) | |
|
| CLIPTextEncoder | SA8775P ADP | SA8775P | QNN | 6.929 ms | 0 - 10 MB | FP16 | NPU | Use Export Script | |
|
| CLIPTextEncoder | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 6.336 ms | 0 - 74 MB | FP16 | NPU | [OpenAI-Clip.tflite](https://huggingface.co./qualcomm/OpenAI-Clip/blob/main/CLIPTextEncoder.tflite) | |
|
| CLIPTextEncoder | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 5.216 ms | 0 - 71 MB | FP16 | NPU | Use Export Script | |
|
| CLIPTextEncoder | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 5.104 ms | 0 - 0 MB | FP16 | NPU | Use Export Script | |
|
|
|
|
|
|
|
|
|
## Installation |
|
|
|
|
|
Install the package via pip: |
|
```bash |
|
pip install "qai-hub-models[openai-clip]" |
|
``` |
|
|
|
|
|
## Configure Qualcomm® AI Hub to run this model on a cloud-hosted device |
|
|
|
Sign-in to [Qualcomm® AI Hub](https://app.aihub.qualcomm.com/) with your |
|
Qualcomm® ID. Once signed in navigate to `Account -> Settings -> API Token`. |
|
|
|
With this API token, you can configure your client to run models on the cloud |
|
hosted devices. |
|
```bash |
|
qai-hub configure --api_token API_TOKEN |
|
``` |
|
Navigate to [docs](https://app.aihub.qualcomm.com/docs/) for more information. |
|
|
|
|
|
|
|
## Demo off target |
|
|
|
The package contains a simple end-to-end demo that downloads pre-trained |
|
weights and runs this model on a sample input. |
|
|
|
```bash |
|
python -m qai_hub_models.models.openai_clip.demo |
|
``` |
|
|
|
The above demo runs a reference implementation of pre-processing, model |
|
inference, and post processing. |
|
|
|
**NOTE**: If you want running in a Jupyter Notebook or Google Colab like |
|
environment, please add the following to your cell (instead of the above). |
|
``` |
|
%run -m qai_hub_models.models.openai_clip.demo |
|
``` |
|
|
|
|
|
### Run model on a cloud-hosted device |
|
|
|
In addition to the demo, you can also run the model on a cloud-hosted Qualcomm® |
|
device. This script does the following: |
|
* Performance check on-device on a cloud-hosted device |
|
* Downloads compiled assets that can be deployed on-device for Android. |
|
* Accuracy check between PyTorch and on-device outputs. |
|
|
|
```bash |
|
python -m qai_hub_models.models.openai_clip.export |
|
``` |
|
``` |
|
Profiling Results |
|
------------------------------------------------------------ |
|
CLIPImageEncoder |
|
Device : Samsung Galaxy S23 (13) |
|
Runtime : TFLITE |
|
Estimated inference time (ms) : 34.3 |
|
Estimated peak memory usage (MB): [0, 54] |
|
Total # Ops : 659 |
|
Compute Unit(s) : NPU (659 ops) |
|
|
|
------------------------------------------------------------ |
|
CLIPTextEncoder |
|
Device : Samsung Galaxy S23 (13) |
|
Runtime : TFLITE |
|
Estimated inference time (ms) : 5.7 |
|
Estimated peak memory usage (MB): [0, 24] |
|
Total # Ops : 660 |
|
Compute Unit(s) : NPU (658 ops) CPU (2 ops) |
|
``` |
|
|
|
|
|
## How does this work? |
|
|
|
This [export script](https://aihub.qualcomm.com/models/openai_clip/qai_hub_models/models/OpenAI-Clip/export.py) |
|
leverages [Qualcomm® AI Hub](https://aihub.qualcomm.com/) to optimize, validate, and deploy this model |
|
on-device. Lets go through each step below in detail: |
|
|
|
Step 1: **Compile model for on-device deployment** |
|
|
|
To compile a PyTorch model for on-device deployment, we first trace the model |
|
in memory using the `jit.trace` and then call the `submit_compile_job` API. |
|
|
|
```python |
|
import torch |
|
|
|
import qai_hub as hub |
|
from qai_hub_models.models.openai_clip import Model |
|
|
|
# Load the model |
|
model = Model.from_pretrained() |
|
image_encoder_model = model.image_encoder |
|
text_encoder_model = model.text_encoder |
|
|
|
# Device |
|
device = hub.Device("Samsung Galaxy S23") |
|
|
|
# Trace model |
|
image_encoder_input_shape = image_encoder_model.get_input_spec() |
|
image_encoder_sample_inputs = image_encoder_model.sample_inputs() |
|
|
|
traced_image_encoder_model = torch.jit.trace(image_encoder_model, [torch.tensor(data[0]) for _, data in image_encoder_sample_inputs.items()]) |
|
|
|
# Compile model on a specific device |
|
image_encoder_compile_job = hub.submit_compile_job( |
|
model=traced_image_encoder_model , |
|
device=device, |
|
input_specs=image_encoder_model.get_input_spec(), |
|
) |
|
|
|
# Get target model to run on-device |
|
image_encoder_target_model = image_encoder_compile_job.get_target_model() |
|
# Trace model |
|
text_encoder_input_shape = text_encoder_model.get_input_spec() |
|
text_encoder_sample_inputs = text_encoder_model.sample_inputs() |
|
|
|
traced_text_encoder_model = torch.jit.trace(text_encoder_model, [torch.tensor(data[0]) for _, data in text_encoder_sample_inputs.items()]) |
|
|
|
# Compile model on a specific device |
|
text_encoder_compile_job = hub.submit_compile_job( |
|
model=traced_text_encoder_model , |
|
device=device, |
|
input_specs=text_encoder_model.get_input_spec(), |
|
) |
|
|
|
# Get target model to run on-device |
|
text_encoder_target_model = text_encoder_compile_job.get_target_model() |
|
|
|
``` |
|
|
|
|
|
Step 2: **Performance profiling on cloud-hosted device** |
|
|
|
After compiling models from step 1. Models can be profiled model on-device using the |
|
`target_model`. Note that this scripts runs the model on a device automatically |
|
provisioned in the cloud. Once the job is submitted, you can navigate to a |
|
provided job URL to view a variety of on-device performance metrics. |
|
```python |
|
image_encoder_profile_job = hub.submit_profile_job( |
|
model=image_encoder_target_model, |
|
device=device, |
|
) |
|
text_encoder_profile_job = hub.submit_profile_job( |
|
model=text_encoder_target_model, |
|
device=device, |
|
) |
|
|
|
``` |
|
|
|
Step 3: **Verify on-device accuracy** |
|
|
|
To verify the accuracy of the model on-device, you can run on-device inference |
|
on sample input data on the same cloud hosted device. |
|
```python |
|
image_encoder_input_data = image_encoder_model.sample_inputs() |
|
image_encoder_inference_job = hub.submit_inference_job( |
|
model=image_encoder_target_model, |
|
device=device, |
|
inputs=image_encoder_input_data, |
|
) |
|
image_encoder_inference_job.download_output_data() |
|
text_encoder_input_data = text_encoder_model.sample_inputs() |
|
text_encoder_inference_job = hub.submit_inference_job( |
|
model=text_encoder_target_model, |
|
device=device, |
|
inputs=text_encoder_input_data, |
|
) |
|
text_encoder_inference_job.download_output_data() |
|
|
|
``` |
|
With the output of the model, you can compute like PSNR, relative errors or |
|
spot check the output with expected output. |
|
|
|
**Note**: This on-device profiling and inference requires access to Qualcomm® |
|
AI Hub. [Sign up for access](https://myaccount.qualcomm.com/signup). |
|
|
|
|
|
|
|
|
|
## Deploying compiled model to Android |
|
|
|
|
|
The models can be deployed using multiple runtimes: |
|
- TensorFlow Lite (`.tflite` export): [This |
|
tutorial](https://www.tensorflow.org/lite/android/quickstart) provides a |
|
guide to deploy the .tflite model in an Android application. |
|
|
|
|
|
- QNN (`.so` export ): This [sample |
|
app](https://docs.qualcomm.com/bundle/publicresource/topics/80-63442-50/sample_app.html) |
|
provides instructions on how to use the `.so` shared library in an Android application. |
|
|
|
|
|
## View on Qualcomm® AI Hub |
|
Get more details on OpenAI-Clip's performance across various devices [here](https://aihub.qualcomm.com/models/openai_clip). |
|
Explore all available models on [Qualcomm® AI Hub](https://aihub.qualcomm.com/) |
|
|
|
|
|
## License |
|
* The license for the original implementation of OpenAI-Clip can be found |
|
[here](https://github.com/openai/CLIP/blob/main/LICENSE). |
|
* The license for the compiled assets for on-device deployment can be found [here](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/Qualcomm+AI+Hub+Proprietary+License.pdf) |
|
|
|
|
|
|
|
## References |
|
* [Learning Transferable Visual Models From Natural Language Supervision](https://arxiv.org/abs/2103.00020) |
|
* [Source Model Implementation](https://github.com/openai/CLIP/) |
|
|
|
|
|
|
|
## Community |
|
* Join [our AI Hub Slack community](https://aihub.qualcomm.com/community/slack) to collaborate, post questions and learn more about on-device AI. |
|
* For questions or feedback please [reach out to us](mailto:[email protected]). |
|
|
|
|
|
|
