Fastllm: A LLM inference library that runs DeepSeek-V4 with 10GB VRAM

 (github.com)

3 points | by nogajun 4 hours ago

0 comments