Unsloth Supervised Fine-Tuning
· Batch Import
Description
Supervised fine-tuning toolkit for instruction-following models using SFTTrainer. Supports chat template mapping, multi-turn conversation synthesis, and standardized dataset formatting (ShareGPT, Alpaca, OpenAI) with optimized processing.
Repository
https://github.com/majiayu000/claude-skill-registry/tree/main/skills/data/unsloth-sft
View on GitHub