Issue 2026-10-12 · Launches · 模型 · 开源

airllm

airllm

AirLLM 70B inference with a single 4GB GPU.

GitHub Trending2 d ago
Read original ↗