HF上，DeepSeek开源OCR模型⬆️今天，DeepSeek开源了最新的模型：DeepSeek-OCR。省流：模型仅3B，单张A100-40G卡每天可跑20万页的LLM/VLM训练数据。更详细来说：DeepSeek提出了一种新的研究——上下文光学 ...

北京时间10月13日，AI大神、OpenAI创始成员之一Andrej Karpathy在GitHub上开源了他的nanochat项目，短短一天时间就收获了上万Star！

Host Keith Shaw and his expert guests discuss the latest technology news and trends happening in the industry. Watch new episodes twice each week or listen to the podcast here. In today’s 2-Minute ...