search

huggingface / alignment-handbook

Robust recipes to align language models with human and AI preferences
https://huggingface.co/HuggingFaceH4
Apache License 2.0
4.2k stars 357 forks source link

Clean deprecated max samples arguments #89

Closed kirill-fedyanin closed 6 months ago

kirill-fedyanin commented 6 months ago

@lewtun, I've removed max_train_samples and max_eval_samples, following the discussion in https://github.com/huggingface/alignment-handbook/pull/80