Commits

Branch selector

 main

User selector

All users

All time

Commit history

Commits on Apr 29, 2026

• Support for a new Granite-Speech-Plus model (#45695)

  Show description for a8f43ec
  

  3 peopleauthored
  a8f43ec

  Copy full SHA for a8f43ec

• [serve] cb error (#45691)

  Show description for c641d13
  SunMarc

  authored
  c641d13

  Copy full SHA for c641d13

• Fix trust_remote_code local cache collisions for local models (#45632) (#45642)

  Show description for 727741f
  Jeevang1-epic

  authored
  727741f

  Copy full SHA for 727741f

• Llama3 video fix (#45040)

  Show description for d1969ba
  

  sywangyi

  and

  zucchini-nlp

  authored
  d1969ba

  Copy full SHA for d1969ba

• [Fix Phi4 test] Fall back to model config for image processor when `trust_remote_code` is False (#45692)

  Show description for db5786e
  yonigozlan

  authored
  db5786e

  Copy full SHA for db5786e

• Fix custom-module copies inheriting read-only permissions (#45686)

  Show description for 1ca0be5
  nurpax

  authored
  1ca0be5

  Copy full SHA for 1ca0be5

• Python code in model docs (#45608)

  Show description for 223fe52
  

  zucchini-nlp

  and

  stevhliu

  authored
  223fe52

  Copy full SHA for 223fe52

• fix failed test cases for blt model (#45596)

  Show description for 9120f5e
  

  kaixuanliu

  and

  ydshieh

  authored
  9120f5e

  Copy full SHA for 9120f5e

• chore(typing): add ty type checking for 3 pipeline files (#45667)

  Show description for 74d576b
  

  moonbogi

  and

  tarekziade

  authored
  74d576b

  Copy full SHA for 74d576b

Commits on Apr 28, 2026

• Add image processors refactor to v5 migration guide (#45556)

  Show description for a374d99
  yonigozlan

  authored
  a374d99

  Copy full SHA for a374d99

• [docs] modular transformers (#45327)

  Show description for 2ed52b1
  stevhliu

  authored
  2ed52b1

  Copy full SHA for 2ed52b1

• change got reverted (#45680)

  itazap

  authored
  b2cb2c3

  Copy full SHA for b2cb2c3

• [docs] dtype (#45659)

  Show description for f3d2b68
  stevhliu

  authored
  f3d2b68

  Copy full SHA for f3d2b68

• fixing more typos (#45689)

  Show description for 93bed93
  

  3 peopleauthored
  93bed93

  Copy full SHA for 93bed93

• [docs] cb memory management (#45587)

  Show description for a29e967
  stevhliu

  authored
  a29e967

  Copy full SHA for a29e967

• [docs] cpu offloading (#45660)

  Show description for 2d01799
  stevhliu

  authored
  2d01799

  Copy full SHA for 2d01799

• docs(README_zh-hans): clarify conditions for not using Transformers (#45688)

  Show description for 4aba716
  GuaiZai233

  authored
  4aba716

  Copy full SHA for 4aba716

• fix padding side issue for fast_vlm tests (#45592)

  Show description for f4fc6d0
  kaixuanliu

  authored
  f4fc6d0

  Copy full SHA for f4fc6d0

• Fix `x_clip`: 8 failed test cases (#45394)

  Show description for 5d7ff43
  kaixuanliu

  authored
  5d7ff43

  Copy full SHA for 5d7ff43

• zero_shot_object_detection ValueError fix for python 3.13 (#45669)

  AnkitAhlawat7742

  authored
  aadc005

  Copy full SHA for aadc005

• Fix pageable H2D copies in Gated DeltaNet PyTorch fallback (#45665)

  Show description for ca72aa0
  ruixiang63

  authored
  ca72aa0

  Copy full SHA for ca72aa0

• Fix UnboundLocalError in shard_and_distribute_module for replicated parameters (#45675)

  Show description for 4287660
  

  Abdennacer-Badaoui

  and

  3outeille

  authored
  4287660

  Copy full SHA for 4287660

• No serving in quality docker image (#45677)

  Show description for 1579083
  ydshieh

  authored
  1579083

  Copy full SHA for 1579083

• Laguna XS.2 implementation (#45673)

  joerowell

  authored
  2df2326

  Copy full SHA for 2df2326

• [MistralCommonBackend] Soften validation mode and apply_chat_template arguments check (#45628)

  Show description for 3146914
  juliendenize

  authored
  3146914

  Copy full SHA for 3146914

Commits on Apr 27, 2026

• Fix `NameError: PeftConfigLike` triggered by `PreTrainedModel.__init_subclass__` (#45658)

  Show description for 61461a7
  qgallouedec

  authored
  61461a7

  Copy full SHA for 61461a7

• Fix cross-attention cache layer type for T5Gemma2 long inputs (#45540)

  Show description for 63378d9
  

  Beichen-Ma

  and

  vasqu

  authored
  63378d9

  Copy full SHA for 63378d9

• chore(typing): added modeling_utils to ty (#45425)

  Show description for 133d68e
  

  tarekziade

  and

  vasqu

  authored
  133d68e

  Copy full SHA for 133d68e

• model: Add DEIMv2 to Transformers (#44339)

  Show description for 739b46c
  

  harshaljanjani

  and

  vasqu

  authored
  739b46c

  Copy full SHA for 739b46c

• [Qwen3.5] Fix GDN linear attention multi-token cached forward (#45513)

  Show description for f53ca05
  kashif

  authored
  f53ca05

  Copy full SHA for f53ca05

• [gemma4] infer from config instead of hardcoding (#45606)

  Show description for 5d24d8c
  

  3 peopleauthored
  5d24d8c

  Copy full SHA for 5d24d8c

• Update quants tests (#45480)

  Show description for e651c68
  SunMarc

  authored
  e651c68

  Copy full SHA for e651c68

• Fix GraniteMoeHybrid _update_mamba_mask crash on attention-only models (#45514)

  Show description for 617e752
  

  3 peopleauthored
  617e752

  Copy full SHA for 617e752

• 🔴🔴🔴 fix: skip `clean_up_tokenization` for BPE tokenizers in `PreTrainedTokenizerFast` (#44915)

  Show description for bbb51c8
  maxsloef-goodfire

  authored
  bbb51c8

  Copy full SHA for bbb51c8

• Fix colmodernvbert tests (#45652)

  Show description for 7fdf09d
  Cyrilvallez

  authored
  7fdf09d

  Copy full SHA for 7fdf09d