저장소
datalab-to/indra
1M param OCR model - runs on CPU, 15x faster than tesseract, close to 1B+ param models in accuracy
- Stars
- ★ 1.3천
- Forks
- 131
- Issues
- 17
- Updated
- 10월 22일
- Language
- Python
- License
- Apache-2.0
#ai#cpu#document-parsing#ocr#onnx#pdf
저장소
1M param OCR model - runs on CPU, 15x faster than tesseract, close to 1B+ param models in accuracy