huggingface/transformers

Pull Requests Commits

Merge branch 'main' into fix-cb-fail-fast

qgallouedec committed 47 days ago

Verified ff9926a7

Fix build_2d_sinusoidal_position_embedding on MPS (#47897)

guarin committed 47 days ago

Verified 562cfd94

Refactor audio buffer creation to use NumPy (#48295)

gsaon committed 47 days ago

Verified efdc97d8

Merge branch 'main' into fix-cb-fail-fast

qgallouedec committed 47 days ago

Verified d0d491da

[docs] Kernel supported models (#48258)

stevhliu committed 47 days ago

Verified ea1410d7

[Gemma4] Investigate flaky test_generation_beyond_sliding_window_1_eager (#48236)

ydshieh committed 47 days ago

Verified 50adfa19

Let gradient checkpointing skip layers with every_n_layers (#48200)

qgallouedec committed 47 days ago

Verified 2b4e2b3e

Disable daily nightly CI (#48292)

remi-or committed 47 days ago

Verified fe1dfb38

Pipeline parallel naive inference (#47289)

3outeille committed 47 days ago

Verified b0c8d427

eustlb committed 47 days ago

Verified da7234ac

CI: fix muse OOMs (#48284)

tarekziade committed 47 days ago

Verified f2c9d590

Fix `BayesianDetectorModel.from_pretrained()` by calling `post_init()` (#48254)

woojinpaik committed 47 days ago

Verified 60f2b9fd

[Fix] Avoid duplicating tests in CI (#48287)

remi-or committed 47 days ago

Verified a548f269

[`GDN`] Fix recurrent FLA fallback (#48266)

vasqu committed 47 days ago

Verified 93fbc550

Fix `tie_word_embeddings` not lifted from `text_config` for some VLM configs (BC regression) (#45857)

qgallouedec committed 47 days ago

Verified cb1e460f

replace xpu-smi subprocess call in benchmark_v2 (#48083)

kaixuanliu committed 47 days ago

Verified b3d7e8c9

ignore mlinter ci file (#48267)

tarekziade committed 47 days ago

Verified 9bb5f6a8

Fix dtype mismatch in grouped_mm_fallback for LoRA training on Mamba+… (#47933)

adh-aakriti committed 47 days ago

Verified 4ad91995

Fail fast when the continuous batching background thread has died

qgallouedec committed 47 days ago

5e2dbfc8

Compute MoE load-balancing loss per layer to avoid giant one-hot materialization −99.7% @ 128k (#48131)

qgallouedec committed 48 days ago

Verified a3536326

deterministic layer_types buffer registration in multiple models (#48162)

mowoe committed 48 days ago

Verified a530903e

Fix nemotron_h save_pretrained emitting singular backbone.embedding.weight (#48075)

yuekaizhang committed 48 days ago

Verified 9916b467

`force_accelerate_hooks` should not hide the signature it wraps (#48156)

SunMarc committed 48 days ago

Verified 89a72731

fix(data_collator): align TokenClassification numpy_call with torch_call (#48212)

Xayar145 committed 48 days ago

Verified 2745e774

[Optimization] Avoid redundant token length initialization loop in StopStringCriteria (#48195)

arnavahire19 committed 48 days ago

Verified b830f5a0

[docs] Cache crop (#47950)

stevhliu committed 48 days ago

Verified 602f674c

Fix a typo in a use of a local variable field_ in a test (#48184)

AleksMat committed 48 days ago

Verified 195734e4

doc: Fix typo in SigLIP2 Flash Attention code example (#48197)

VimalN2005 committed 48 days ago

Verified fe314d40

[serge] Fix 2 integration tests regressed by commit 16780c86b20a (PR #47622) (#48134)

sergereview[bot] committed 48 days ago

Verified 17044fa5

[xcodec2] Fix flex attention and flash dispatch tests (#48244)

jiqing-feng committed 48 days ago

Verified c7cf04b1

Older