Hacker News new | ask | show | jobs
Fastllm: A LLM inference library that runs DeepSeek-V4 with 10GB VRAM (github.com)
3 points by nogajun 32 days ago