The DeepSeek team open-sources a new model, DeepSeek-OCR
By:Cointime
DeepSeek-AI team released the paper "DeepSeek-OCR: Contexts Optical Compression", proposing a new method for compressing long text context using visual modalities. The Hugging Face page shows that the model has 3 billion parameters.
0
0
Disclaimer: The content of this article solely reflects the author's opinion and does not represent the platform in any capacity. This article is not intended to serve as a reference for making investment decisions.
You may also like
FTC deletes posts from Lina Khan's tenure discussing AI dangers and open source
Bitget-RWA•2025/10/20 23:57
Anthropic introduces Claude Code for online use
Bitget-RWA•2025/10/20 23:57
Trump DOE announces the withdrawal of more than $700 million in manufacturing funding
Bitget-RWA•2025/10/20 23:57
US and Australia reach $3B agreement on essential minerals
Bitget-RWA•2025/10/20 23:57