main
[style] Rework ruff rules and update all files (#43144)
Fix local dataset loading priority in run_image_classification_no_tra… (#44199)
fix: Better weight decay exclusion in `run_*_no‑trainer.py` examples (#43947)
Fix weight decay exclusions in `run_*_no‑trainer.py` examples (#42769)
Remove more deprecated objects/args (#43195)
Train sampler unification (#43138)
rm slow tokenizers (#40936)
Remove Neptune integration references and deprecate `NeptuneCallback` (#42666)
Add torchcodec in docstrings/tests for `datasets` 4.0 (#39156)
improve dev setup comments and hints (#28495)
⚠️⚠️ Use `dtype` instead of `torch_dtype` everywhere! (#39782)
[CB] [Minor] Add parameter to tune default compile level (#46533)
enable cpu paged cache (#42869)
deprecate `overwrite_output_dir` (#41323)
Add slow to some examples tests (#42164)
[refactor] Serving into proper modules (#44796)
remove `tpu_num_cores` (#41383)