Commits

Branch selector

 main

User selector

All users

All time

Commit History

Commits on Apr 6, 2026

• Add `hasattr(torch.backends.cudnn, "conv")` to `conftest.py` (#45263)

  Show description for b9f0fbf
  ydshieh

  authored
  b9f0fbf

  Copy full SHA for b9f0fbf

• Fix `SmolVLM` video processor `resize` using wrong interpolation after backend refactor (#45258)

  Show description for 182f20c
  ydshieh

  authored
  182f20c

  Copy full SHA for 182f20c

• Fix `Qwen2IntegrationTest` (#45268)

  Show description for 02c38f7
  ydshieh

  authored
  02c38f7

  Copy full SHA for 02c38f7

• doc: fix TokenizersBackend.convert_to_native_format docstring (#45262)

  lowzhao

  authored
  5182cf3

  Copy full SHA for 5182cf3

• empty (#45261)

  Show description for d3c7a19
  ydshieh

  authored
  d3c7a19

  Copy full SHA for d3c7a19

Commits on Apr 5, 2026

• Fix unexpected TF32 being enabled in testing (#45252)

  Show description for 374d44d
  ydshieh

  authored
  374d44d

  Copy full SHA for 374d44d

• Fix tf32 issue: set `torch.backends.cudnn.conv.fp32_precision` explicitly. (#45248)

  Show description for 794d65f
  ydshieh

  authored
  794d65f

  Copy full SHA for 794d65f

Commits on Apr 4, 2026

• Nvidia CI with `torch 2.11` (#45243)

  Show description for 499ef1d
  ydshieh

  authored
  499ef1d

  Copy full SHA for 499ef1d

• Update tiny model creation script (#45241)

  Show description for eb981ae
  

  ydshieh

  and

  claude

  authored
  eb981ae

  Copy full SHA for eb981ae

• Update `get_test_info.py` (related to tiny model creation) (#45238)

  Show description for 4b8fdf4
  ydshieh

  authored
  4b8fdf4

  Copy full SHA for 4b8fdf4

Commits on Apr 3, 2026

• More fix for tiny model creation (#45228)

  Show description for e1b80de
  ydshieh

  authored
  e1b80de

  Copy full SHA for e1b80de

• remove unnecessary entries in some auto model mappings (#45224)

  Show description for 09a3e08
  ydshieh

  authored
  09a3e08

  Copy full SHA for 09a3e08

• fix: hf-doc-builder insallation was failing (#45225)

  Show description for df557a2
  tarekziade

  authored
  df557a2

  Copy full SHA for df557a2

• [CB] Add per-request logits processors (#45026)

  Show description for 82a73ad
  remi-or

  authored
  82a73ad

  Copy full SHA for 82a73ad

• [docs] formatting (#45196)

  Show description for db9f18c
  

  stevhliu

  and

  tarekziade

  authored
  db9f18c

  Copy full SHA for db9f18c

• fix `test_register_result_handler` (#45188)

  Show description for 66b6e9f
  SunMarc

  authored
  66b6e9f

  Copy full SHA for 66b6e9f

• [CB] Tweaks to update and minor fixes (#45179)

  Show description for 138f757
  remi-or

  authored
  138f757

  Copy full SHA for 138f757

• Fix pypi release (#45210)

  Show description for e958b56
  ArthurZucker

  authored
  e958b56

  Copy full SHA for e958b56

• update to dev version 5.6.0-dev0

  ArthurZucker

  committed
  c38b2fb

  Copy full SHA for c38b2fb

Commits on Apr 2, 2026

• fix(docs): correct gemma4 docs and examples (#45197)

  Show description for edaac7d
  douglas-reid

  authored
  edaac7d

  Copy full SHA for edaac7d

• Add Turkish (tr) translation for Get Started section (#45158)

  Show description for 3c307e3
  onwp

  authored
  3c307e3

  Copy full SHA for 3c307e3

• [docs] transformers serve (#45174)

  Show description for a4cf3b8
  stevhliu

  authored
  a4cf3b8

  Copy full SHA for a4cf3b8

• casually dropping the most capable open weights on the planet (#45192)

  Show description for 91b1ab1
  

  14 peopleauthored
  91b1ab1

  Copy full SHA for 91b1ab1

• Internalise the NomicBERT model (#43067)

  Show description for a594e09
  

  3 peopleauthored
  a594e09

  Copy full SHA for a594e09

• Fix resized LM head weights being overwritten by post_init (#45079)

  Show description for 4932e97
  javierdejesusda

  authored
  4932e97

  Copy full SHA for 4932e97

• [Qwen3.5 MoE] Add _tp_plan to ForConditionalGeneration (#45124)

  Show description for 57e8413
  danielquintas8

  authored
  57e8413

  Copy full SHA for 57e8413

• Fix TypeError: 'NoneType' object is not iterable in GenerationMixin.generate (#45164)

  Show description for b10552e
  albertvillanova

  authored
  b10552e

  Copy full SHA for b10552e

• fix(models): Fix dtype mismatch in SwitchTransformers and TimmWrapperModel (#45074)

  Show description for 423f2a3
  harshaljanjani

  authored
  423f2a3

  Copy full SHA for 423f2a3

• Generalize gemma vision mask to videos (#45185)

  Show description for ade7a05
  zucchini-nlp

  authored
  ade7a05

  Copy full SHA for ade7a05

• [misc] fix qwen35 tests: correct the text model type and skip reverse_mapping (#45173)

  Show description for 5383689
  JJJYmmm

  authored
  5383689

  Copy full SHA for 5383689

• 🔒 Pin GitHub Actions to commit SHAs (#45180)

  Show description for abc417a
  paulinebm

  authored
  abc417a

  Copy full SHA for abc417a

• Use doc-builder runnable example for GLM-ASR (#44277)

  Show description for a118714
  tarekziade

  authored
  a118714

  Copy full SHA for a118714

• CI] Small T5 expectations updated (#45138)

  Show description for bb80310
  Abdennacer-Badaoui

  authored
  bb80310

  Copy full SHA for bb80310

Commits on Apr 1, 2026

• fix: correct type annotations across config classes for @strict validation (#45007)

  Show description for f38d663
  

  Krishnachaitanyakc

  and

  zucchini-nlp

  authored
  f38d663

  Copy full SHA for f38d663

• Fix explicit local code resolution for tokenizers and image processors (#45169)

  Show description for 43ed69f
  hmellor

  authored
  43ed69f

  Copy full SHA for 43ed69f