Skip to content

Pull requests: huggingface/transformers

New pull request New
1,232 Open 24,310 Closed
1,232 Open 24,310 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

[MOE] MoE routing capture and replay support
#44925 opened Mar 22, 2026 by kashif Loading…
5 tasks
1
Continuous batching paged attention thread safety
#44924 opened Mar 22, 2026 by Qubitium Loading…
2 of 5 tasks
1
fix: avoid unconditional model_info call in _patch_mistral_regex
#44923 opened Mar 22, 2026 by prakhar-agarwal • Draft
fix: pop output_* flags from kwargs in capture_outputs to prevent submodule leakage
#44922 opened Mar 22, 2026 by s-zx Loading…
1
1
fix: use shape index access in compute_3d_position_ids for Qwen VL models
#44921 opened Mar 22, 2026 by s-zx Loading…
1
1
fix: propagate num_labels/id2label to text_config in Qwen3_5Config
#44920 opened Mar 22, 2026 by s-zx Loading…
1
1
fix: handle ragged batch inputs in Qwen2_5_VLProcessor mm_token_type_ids computation
#44919 opened Mar 21, 2026 by s-zx Loading…
1
1
fix(gpt-neox): preserve rotary_pct across save/load cycle
#44917 opened Mar 21, 2026 by s-zx Loading…
1
1
fix(deberta-v2): move "Copied from" comments above @torch.jit.script for Python 3.13 compat
#44916 opened Mar 21, 2026 by s-zx Loading…
1
1
fix: skip clean_up_tokenization for BPE tokenizers in PreTrainedTokenizerFast
#44915 opened Mar 21, 2026 by maxsloef-goodfire Loading…
3 of 5 tasks
2
fix: set clean_up_tokenization_spaces=False in Llama 3 tokenizer conversion
#44914 opened Mar 21, 2026 by maxsloef-goodfire Loading…
2 of 5 tasks
2
1
Fix flash attention crash with 3D position_ids (Qwen3.5)
#44911 opened Mar 21, 2026 by ouroborosscr Loading…
2 of 5 tasks
1
2
Fix: Update optimization.py
#44909 opened Mar 21, 2026 by anshuS1310 Loading…
1
Remove unnecessary expand_as in get_placeholder_mask across VLMs
#44907 opened Mar 21, 2026 by syncdoth Loading…
7 tasks done
1
1
fix(i18n): replace broken relative links to awesome-transformers.md with absolute URLs
#44905 opened Mar 21, 2026 by NicoleRobin Loading…
fix(granite_speech): convert int to float for multiplier fields in text_config
#44904 opened Mar 21, 2026 by vivekvar-dl Loading…
5 tasks done
1
1
Support SizeDict import in get_size_dict
#44903 opened Mar 21, 2026 by yonigozlan Loading…
6
fix: Whisper word timestamp OOB access on trailing replacement char
#44902 opened Mar 20, 2026 by guoyangzhen Loading…
1
1
fix(models): Fix Perceiver interpolate_pos_encoding interpolating to the source size
#44899 opened Mar 20, 2026 by harshaljanjani Loading…
3 of 5 tasks
1
1
[docs] continuous batching
#44896 opened Mar 20, 2026 by stevhliu Loading…
1
Add static FP8 expert support
#44895 opened Mar 20, 2026 by SunMarc Loading…
3
fix processing_utils.py: avoid deepcopying tokenizer in ProcessorMixin to improve performance
#44894 opened Mar 20, 2026 by ydshieh Loading…
2
add StaticLayer.crop() to match DynamicLayer API
#44893 opened Mar 20, 2026 by ai-man-codes Loading…
2 of 5 tasks
1
1
Fix AutoImageProcessor.from_pretrained failing on URL input
#44892 opened Mar 20, 2026 by he-yufeng Loading…
1
[Trainer] add MoERouterHealthCallback Callback
#44891 opened Mar 20, 2026 by kashif Loading…
5 tasks
2
ProTip! Find all pull requests that aren't related to any open issues with -linked:issue.