Fine-tuning a 350M Model for Better Structured Outputs in 100 GRPO Steps
Hugging Face
LFM2.5-350M was fine-tuned with GRPO using TRL for structured-output compliance and then re-evaluated on IFStruct. The run used about 500 samples and 100 training steps, improving the IFStruct pass rate from 22.6% to 29.7%. JSON compliance increased from 18.0% to 31.9% while YAML stayed roughly flat, indicating task-specific fine-tuning shifts the model toward the targeted output formats.
Related stories
Fine-Tuning Tool-Calling LLMs: A Complete Guide Using XYZ-Aquila-SFT and Qwen3
MarkTechPost · 2 months ago ·
48
Fine-Tuning Small Open-Source LLMs to Outperform Large Closed-Source Models by 60% on Specialized Tasks
Together AI · 1 year ago ·
56
Preparing data for supervised fine-tuning Part 1: Formatting and quality
Amazon Web Services · 1 month ago ·
34