Skip to content

Pull requests: huggingface/transformers

New pull request New
1,312 Open 24,975 Closed
1,312 Open 24,975 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

[CB] Hide activation footprint by using the CUDA graph pool
#45911 opened May 12, 2026 by remi-or Collaborator Loading…
@remi-or
2
Automatically inherit properties from children in composite models
#45909 opened May 12, 2026 by Cyrilvallez Member Loading…
2
Fix M-RoPE inv_freq device and meta → to_empty re-init in Qwen3-VL family
#45903 opened May 12, 2026 by jamesbraza Contributor Loading…
1 task done
2
1
[docs] ALMModelTest
#45900 opened May 11, 2026 by stevhliu Member Loading…
1
[docs] decode fast path
#45899 opened May 11, 2026 by stevhliu Member Loading…
1
[fix] Add fatal_error to ContinuousBatchingManager so the serving layer can detect a dead CB worker
#45898 opened May 11, 2026 by qgallouedec Member Loading…
6
Enhance apply_chat_template to support custom field prefilling (reasoning_content, thinking, etc.)
#45896 opened May 11, 2026 by Mamiglia Loading…
3 of 5 tasks
3
Fix undefined 'input' variable
#45895 opened May 11, 2026 by fullyz Contributor Loading…
3 of 6 tasks
2
Fix deepseek v4
#45892 opened May 11, 2026 by ArthurZucker Collaborator Loading…
6 tasks
5
exaone4_5: add XPU expectations
#45890 opened May 11, 2026 by kaixuanliu Contributor Loading…
1
fix(rope): read original_max_position_embeddings from yarn validator's argument
#45887 opened May 11, 2026 by bzantium Contributor Loading…
5
Fix/pe audio video bugs
#45886 opened May 11, 2026 by massimilianoviola Loading…
9
Fix Gemma4 inputs_embeds OOM during per-layer lookup
#45883 opened May 11, 2026 by chirag-gupta-07 Loading…
2
add DeepSeek-V4-Flash-Base support, also add the testcase(dequantize=…
#45880 opened May 11, 2026 by sywangyi Contributor Loading…
13
feat(t5gemma2): add Flash Attention 2 support
#45868 opened May 10, 2026 by AjAyrAo43 Loading…
9
fix: add HF_USE_MLX opt-out for MLX detection
#45863 opened May 9, 2026 by harryfrzz Loading…
5 of 6 tasks
1
[new model] Add Zyphra/ZAYA1-8B
#45862 opened May 9, 2026 by JJJYmmm Contributor Loading…
39
Fix M-RoPE device mismatch in Qwen3VL family under FSDP2 CPU offload
#45861 opened May 9, 2026 by jamesbraza Contributor Loading…
1 task done
1
2
fix(text-generation): use token-level slicing for return_full_text=Fa…
#45860 opened May 9, 2026 by Abineshabee Contributor Loading…
1
7
hy_v3: add XPU expectations
#45858 opened May 9, 2026 by kaixuanliu Contributor Loading…
1
Fix tie_word_embeddings not lifted from text_config for some VLM configs (BC regression)
#45857 opened May 9, 2026 by qgallouedec Member Loading…
4
Fix slow Trainer path with 4D attention mask
#45852 opened May 8, 2026 by qflen Loading…
3 of 5 tasks
1
3
[docs] chat template
#45851 opened May 8, 2026 by stevhliu Member Loading…
1
1
Add Maximal Update Parametrization (μP)
#45848 opened May 8, 2026 by qgallouedec Member Loading…
1
2
Streamable chat parsing
#45847 opened May 8, 2026 by Rocketknight1 Member Loading…
2 of 4 tasks
3
ProTip! Adding no:label will show everything without a label.