FINE-TUNING
5 items
4 posts, 1 tool
OpenAI stops new fine-tuning jobs on January 6, 2027, and every base model it still lets you tune shuts down by January 14. The dates, what keeps serving, where fine-tuning still exists, which providers let you download weights, and a migration checklist.
Fine-tuned coding agents collapse when you switch scaffolds, from 52.6% to 8.4% on SWE-bench Verified. DCAS shows how to train planning into the model.
A $500 GRPO fine-tune of Qwen 3.5 9B beat GPT-5.6 Sol and Opus 4.8 on catalog review, hitting 97% accuracy at a fraction of the cost per listing.
Gemma 4 ships byte-for-byte open weights from Google DeepMind. How developers deploy it locally, fine-tune it, and ship agents on top of it.
Meta's open-weight model family. Llama 4, released April 5, 2025, comes in Scout (109B total, 17B active) and Maverick (400B total, 17B active). Commercial use under the Llama license.

Get Smarter About AI Dev
New tutorials, open-source projects, and deep dives on coding agents - delivered weekly.