Skip to content

Pull requests: huggingface/transformers

New pull request New
1,280 Open 24,672 Closed
1,280 Open 24,672 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

Add EXAONE 4.5 implementations
#45471 opened Apr 16, 2026 by nuxlear Contributor Loading…
3 of 6 tasks
2
skip test_flash_attn_2_can_dispatch_composite_models tests for
#45470 opened Apr 16, 2026 by kaixuanliu Contributor Loading…
1
Fix: propagate interpolate_pos_encoding through Pixio model hierarchy
#45469 opened Apr 16, 2026 by Spectual Loading…
3 of 6 tasks
1
1
Fix MPS SDPA output shape when value head dim differs from query head dim
#45467 opened Apr 16, 2026 by Jah-yee Loading…
1
[docs] contributing
#45465 opened Apr 15, 2026 by stevhliu Member • Draft
1
Fix response api support
#45463 opened Apr 15, 2026 by SunMarc Member Loading…
6
chore(sec): added a handful of security checks
#45462 opened Apr 15, 2026 by tarekziade Collaborator Loading…
5
Remove redundant condition checks in get_image_size method
#45461 opened Apr 15, 2026 by JiauZhang Loading…
3
fix(tokenization): re-raise ImportError to allow RuntimeError/OSError fallback (#45459)
#45460 opened Apr 15, 2026 by cloudyun888 Loading…
1
Allow loading Qwen Thinker 'base' models without generative head
#45457 opened Apr 15, 2026 by tomaarsen Member Loading…
2 of 6 tasks
11
[fix] Make Qwen2_5OmniProcessor warning a lot less noisy via warning_once
#45455 opened Apr 15, 2026 by tomaarsen Member Loading…
2 of 6 tasks
10
Gemma4 training with text-only samples
#45454 opened Apr 15, 2026 by zucchini-nlp Member Loading…
1
13
Draft commit
#45453 opened Apr 15, 2026 by ArthurZucker Collaborator • Draft
2
refactor: replace wildcard imports with explicit imports in model __init__.py files
#45452 opened Apr 15, 2026 by DavidSolanas Loading…
1
2
chore: bump doc-builder SHA for PR upload workflow
#45450 opened Apr 15, 2026 by rtrompier Loading…
1
[loading] Clean way to add/remove full parts in checkpoint names
#45448 opened Apr 15, 2026 by Cyrilvallez Member Loading…
13
[fix] Always early return for non-Mistral models in _patch_mistral_regex
#45444 opened Apr 14, 2026 by tomaarsen Member Loading…
3 of 6 tasks
1
11
Raise 400 on model mismatch when transformers serve is pinned
#45443 opened Apr 14, 2026 by qgallouedec Member Loading…
2
fix(DSV3): parity between native DeepseekV3MoE and remote official implementation
#45441 opened Apr 14, 2026 by casinca Contributor Loading…
4 of 6 tasks
1
4
Add Gemma4ForSequenceClassification
#45438 opened Apr 14, 2026 by Charly21r Contributor Loading…
4 tasks done
1
19
Fix spurious position_ids warnings for at least 40 architectures
#45437 opened Apr 14, 2026 by tomaarsen Member Loading…
2 of 6 tasks
6
Add expert parallelism (EP) support for Qwen3 MoE + fix GroupedGemmParallel for 2D meshes
#45436 opened Apr 14, 2026 by AmineDiro Member Loading…
7 tasks done
4
do not index past decoded chars with special tokens
#45435 opened Apr 14, 2026 by itazap Collaborator Loading…
1
2
SonicMoe
#45433 opened Apr 14, 2026 by IlyasMoutawwakil Member • Draft
6 tasks
1
chore(qa): split pipeline and add type checking
#45432 opened Apr 14, 2026 by tarekziade Collaborator Loading…
@tarekziade
3
ProTip! What’s not been updated in a month: updated:<2026-03-16.