tag

#qwen

1 post

2026

Oct 24, 2026A 4B model, my RAG system, and what actually workedTwo weeks chasing a small model as good as GLM 5 on my own work. Memory inside the model failed; teaching a 4B to use my RAG tools got it within 2 points of the cloud model it runs on.6 min read3 interactivellmsmall-modelsragdistillation