Train sampler unification (#43138)
* add sequential sampling
* train_sampler arg
* Update src/transformers/training_args.py
Co-authored-by: Marc Sun <57196510+SunMarc@users.noreply.github.com>
* updated train_sampler to train_sampling_strategy; updated warnings/docs
* update docs that prev reference group_by_length
* Clean up post-merge: remove duplicate fields and fix model card date
Remove 13 duplicate field definitions in TrainingArguments that existed
in both our old layout and upstream's reorganized sections. Fix
exaone_moe model card date per utils/add_dates.py.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
---------
Co-authored-by: Marc Sun <57196510+SunMarc@users.noreply.github.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
