OtotaO_unsloth-mcp-server
0
Unsloth MCP Server optimizes the fine-tuning of large language models, offering faster speeds and reduced memory usage without compromising quality. It provides a wide range of model support and uses advanced techniques like 4-bit quantization for efficiency.
Unsloth MCP Server
An MCP server for Unsloth, a library designed to optimize the fine-tuning of large language models. Key features include: 2x faster fine-tuning, 80% less memory usage, extended context lengths, and no loss in model performance. Supports models like Llama, Mistral, Phi, and Gemma with 4-bit quantization and provides tools for installation checking, supported models listing, model loading, fine-tuning, text generation, and model exporting.