DPOTrainer

A developer reports that DPO training using Hugging Face's DPOTrainer on a fine-tuned T5 model causes the model to generate repetitive outputs (e.g., 'a a a a a a') and degrade conversational coherenc…

Developer Ammar Hassona released Trainsafe, a TrainerCallback for HuggingFace/TRL that runs behavioral health checks during fine-tuning, catching failures like language drift and output collapse that …