Fix dataset preprocessing cache reuse in DPOTrainer and SFTTrainer . Reprocessing the same raw dataset with…
Fix dataset preprocessing cache reuse in DPOTrainer and SFTTrainer . Reprocessing the same raw dataset with…: a task in HF ML Tasksmith (Harbor dataset). Preserve preprocessing behavior for both plain-text and conversational datasets, including prompt/completion boundaries, preference outputs, EOS…
The task
Preserve preprocessing behavior for both plain-text and conversational datasets, including prompt/completion boundaries, preference outputs, EOS handling, tools, per-example chat-template options, and assistant/completion masks. SFT must continue to honor its selected chat template. VLM tokenization must continue to…
Part of FineEnvs/HF_ML_Tasksmith.