https://github.com/first-coding/vit
This project distills a ViT model into a compact CNN, reducing its size to 1.24MB with minimal accuracy loss. ONNXRuntime with CUDA boosts inference speed, while FastAPI and Docker simplify deployment.
https://github.com/first-coding/vit
docker fastapi image-classification knowledge-distillation model-quantization onnx onnxruntime python vision-transformer
Last synced: about 1 month ago
JSON representation
This project distills a ViT model into a compact CNN, reducing its size to 1.24MB with minimal accuracy loss. ONNXRuntime with CUDA boosts inference speed, while FastAPI and Docker simplify deployment.
- Host: GitHub
- URL: https://github.com/first-coding/vit
- Owner: first-coding
- Created: 2024-05-30T09:39:10.000Z (almost 2 years ago)
- Default Branch: master
- Last Pushed: 2025-05-17T05:00:41.000Z (about 1 year ago)
- Last Synced: 2025-07-30T04:00:51.980Z (10 months ago)
- Topics: docker, fastapi, image-classification, knowledge-distillation, model-quantization, onnx, onnxruntime, python, vision-transformer
- Language: Python
- Homepage:
- Size: 632 MB
- Stars: 2
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md