X · @ollama
· X / Twitter
RT Burke Holland: Qwen3-Coder 30B running locally in the @GitHub Copilot app (@ollama) spitting ~85 tokens per second on 96GB VRAM. It's not Sol, but …
RT Burke HollandQwen3-Coder 30B running locally in the @GitHub Copilot app (@ollama) spitting ~85 tokens per second on 96GB VRAM. It's not Sol, but we're getting somewhere.