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

3 pointsposted 10 hours ago
by nogajun
(github.com)

No comments yet