arXiv NLP
techCenter
Jina-OCR-v1: Efficient Document Parsing with Speculative Decoding and Dense Verifiable Rewardstranslating…
arXiv:2609.03181v1 Announce Type: new
Abstract: We present Jina-OCR-v1, an end-to-end document parsing model built to serve on low-budget GPUs. It combines the compressed-vision encoder and the 3B mixture-of-experts decoder of DeepSeek-OCR, which activates about 570M parameters…
Keywords#Document Parsing#Jina-OCR-v1#Decoding#Dense#Dense Verifiable
Comments
Sign in to join the discussion.
Loading comments…