Установить Turboquant

MCP server for LLM quantization. Compress any HuggingFace model to GGUF, GPTQ, or AWQ format. 6 tools: info, check, recommend, quantize, evaluate, push. Self-co