"author:shimizu06" matching MCP servers:

Enables users to describe an LLM fine-tuning job in natural language so the server estimates required VRAM/FLOPs and searches live GPU pricing across a dozen cloud providers. Returns ranked cheapest, fastest, and balanced hardware recommendations with each result's data source clearly identified.