Ai-Agents

My Local LLM Scored 6/6. It Was Wrong Every Time.

Six months of trying to make a 1.2B model useful, and the measurement mistakes I made along the way.