Update README.md

Upload README.md

Upload training_metadata.json with huggingface_hub

DPO adapter: Qwen/Qwen2.5-3B-Instruct, beta=0.5, 12 pairs, over_mask_heldout=0.0%, t8_overmask=0.0%, t8_exact_match=57.1%

Upload training_metadata.json with huggingface_hub

Upload model trained with Unsloth

DPO adapter: Qwen/Qwen2.5-3B-Instruct, beta=0.5, 12 pairs, over_mask_heldout=0.0%, t8_overmask=0.0%

initial commit