Skip to content

Pull requests: huggingface/transformers

New pull request New
1,022 Open 22,219 Closed
1,022 Open 22,219 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

Fix logic in setting self.fsdp when it is False
#41974 opened Nov 2, 2025 by roychan Loading…
Fix import error with huggingface_hub v1.0.0+
#41973 opened Nov 2, 2025 by jameslovespancakes Loading…
1
1
fix: improve visibility of ValueError root causes in model config loading
#41972 opened Nov 1, 2025 by scottzh8 Loading…
add support for saving encoder only so any parakeet model can be loaded for inference
#41969 opened Oct 31, 2025 by nithinraok Loading…
5 tasks
1
test tensor parallel: make tests for dense model more robust
#41968 opened Oct 31, 2025 by 3outeille Loading…
1
feat: RoPE-related typing improvements
#41967 opened Oct 31, 2025 by RyanMullins • Draft
4 of 5 tasks
1
1
Fix GPT-2 Flash Attention 2 generation with left-padding
#41966 opened Oct 31, 2025 by Abdennacer-Badaoui Loading…
3
Fix: add safe helper for load_gguf_checkpoint in PreTrainedConfig
#41965 opened Oct 31, 2025 by AnkanJU Loading…
5 tasks
fix: dict[RopeParameters] to dict[str, RopeParameters]
#41963 opened Oct 31, 2025 by RyanMullins Loading…
3 of 5 tasks
1
Patch Mistral Common Tokenizer 2: fix add_special_tokens
#41962 opened Oct 31, 2025 by juliendenize Loading…
2 of 5 tasks
[XPU] Add flash_attn2 support for XPU
#41956 opened Oct 30, 2025 by YangKai0616 • Draft
4
Fix max_pixels/min_pixels ignored in Qwen2VLImageProcessorFast.from_pretrained
#41954 opened Oct 30, 2025 by ReinforcedKnowledge Loading…
1
Merge Text2Text into TextGenerationPipeline
#41953 opened Oct 30, 2025 by Rocketknight1 • Draft
Fixes wrong epoch count in trainer under some conditions
#41952 opened Oct 30, 2025 by i3hz Loading…
2 of 5 tasks
1
Fixed wrong padding value in OWLv2
#41938 opened Oct 29, 2025 by gjamesgoenawan Loading…
2
Refactor: Replace _default_log_level with DEFAULT_LOG_LEVEL constant
#41937 opened Oct 29, 2025 by Pranavi125 Loading…
1
Fix: add missing SAFE_WEIGHTS_INDEX_NAME to __all__ in constants.py
#41936 opened Oct 29, 2025 by Pranavi125 Loading…
2
Fix: Handle missing safetensors gracefully to prevent import errors
#41932 opened Oct 29, 2025 by Pranavi125 Loading…
2
fix 3 failed test cases for video_llama_3 model on Intel XPU
#41931 opened Oct 29, 2025 by kaixuanliu Loading…
20
handle inputs from Siglip/Siglip2 non-automapped encoder layers
#41930 opened Oct 29, 2025 by molbap Loading…
1
29
fix: add clear error message when mistral-common is missing for AutoTokenizer loading Voxtral
#41928 opened Oct 29, 2025 by junjunjd Loading…
5 tasks
4
[deepspeed tests fixes]
#41925 opened Oct 28, 2025 by stas00 Loading…
5
fix tensor device placement issue of 2 UT cases
#41921 opened Oct 28, 2025 by yao-matrix Loading…
2
feat(ci): add continuous batching to benchmarks run-benchmark
#41916 opened Oct 28, 2025 by McPatate • Draft
1
Fix Break change of AWQ FusedModules due to Attention Refactor
#41909 opened Oct 28, 2025 by fanqiNO1 Loading…
2 of 5 tasks
1
ProTip! Filter pull requests by the default branch with base:main.