hugging-face-model-trainer

Trains and fine-tunes language models using TRL (Transformer Reinforcement Learning) on Hugging Face Jobs infrastructure, supporting SFT, DPO, GRPO, reward modeling, and GGUF conversion for local deployment.