-
Notifications
You must be signed in to change notification settings - Fork 33k
Pull requests: huggingface/transformers
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
Add image processors refactor to v5 migration guide
#45556
opened Apr 21, 2026 by
yonigozlan
Member
Loading…
perf(
gemma3): Avoid recomputing rotary_emb for each layer
#45555
opened Apr 21, 2026 by
casinca
Contributor
Loading…
3 of 6 tasks
Add ForSequenceClassification heads for the OLMo family
#45551
opened Apr 21, 2026 by
earino
Loading…
5 of 6 tasks
Add runner selection for mi325 GPU type
#45550
opened Apr 21, 2026 by
glegendre01
Contributor
•
Draft
6 tasks
fix: apply channel averaging correctly in audio feature extractors
#45549
opened Apr 21, 2026 by
jonghwanhyeon
Contributor
Loading…
3 of 6 tasks
Fix EP + DeepSpeed ZeRO-3 loading via accelerate launch
#45548
opened Apr 21, 2026 by
AmineDiro
Member
Loading…
2 tasks done
Add disable_mmap kwarg to from_pretrained with hf-mount auto-detection
#45547
opened Apr 21, 2026 by
rtrompier
Contributor
Loading…
feat: Add GGUF loading support for Llama 4 (text)
#45546
opened Apr 21, 2026 by
garybadwal
Loading…
4 of 6 tasks
Fix local_files_only tokenizer fallback when tokenizer files are missing (Issue 45538)
#45541
opened Apr 21, 2026 by
Brianzhengca
Loading…
4 of 7 tasks
Fix cross-attention cache layer type for T5Gemma2 long inputs
#45540
opened Apr 21, 2026 by
Beichen-Ma
Loading…
4 of 6 tasks
[modular] Fix modular logic broken in #45045
#45539
opened Apr 21, 2026 by
Cyrilvallez
Member
Loading…
[CB] Changes for long generation
#45530
opened Apr 20, 2026 by
remi-or
Collaborator
Loading…
4 tasks done
utils: handle flash_attn missing from importlib packages_distributions without crashing
#45524
opened Apr 20, 2026 by
SAY-5
Loading…
Fix Seq2SeqLM ExecuTorch export: add encoder_attention_mask to decoder and use static encoder shapes
#45523
opened Apr 20, 2026 by
duyhv-qualgo
Loading…
3 tasks
Fix GraniteMoeHybrid _update_mamba_mask crash on attention-only models
#45514
opened Apr 19, 2026 by
tianhaocui
Loading…
[Qwen3.5] Fix Qwen3.5 linear attention multi-token cached forward
#45513
opened Apr 19, 2026 by
kashif
Contributor
Loading…
6 tasks
Add full GGUF loading support for GPT‑OSS (fixes #43366, supersedes #43757) latest
#45506
opened Apr 18, 2026 by
sirzechs66
Loading…
5 of 6 tasks
Previous Next
ProTip!
Adding no:label will show everything without a label.