Skip to content

Pull requests: huggingface/transformers

New pull request New
681 Open 19,197 Closed
681 Open 19,197 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
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

Remove torchvision requirement from AutoImageProcessor
#37457 opened Apr 11, 2025 by LysandreJik Loading…
2
Update check_modular_conversion
#37456 opened Apr 11, 2025 by qubvel • Draft
1
feat: Add gradient testing for Flash Attention 2
#37453 opened Apr 11, 2025 by crStiv • Draft
2
Set should_evaluate=eval_on_start on train start
#37450 opened Apr 11, 2025 by I-l-l-I Loading…
1
[Gemma3] compile ✨
#37447 opened Apr 11, 2025 by gante Loading…
2 tasks done
12
remove _run_third_party_device_tests
#37445 opened Apr 11, 2025 by jiqing-feng Loading…
1
Fix Aria tests
#37444 opened Apr 11, 2025 by jiqing-feng Loading…
3
[bug] deprecated deta load_cuda_kernel, MultiScaleDeformableAttention
#37443 opened Apr 11, 2025 by chagmgang Loading…
5 tasks
2
Implemented update function in cache_utils.py, with a test file test_cache_utils.py
#37442 opened Apr 11, 2025 by ailunc • Draft
4 of 5 tasks
2
Update quantization docs
#37439 opened Apr 11, 2025 by DerekLiu35 Loading…
11
fix issue that some example with no trainer use accelerator.end_train…
#37435 opened Apr 10, 2025 by we1559 Loading…
2 of 5 tasks
1
2
allow accelerate prepare for torch compiled models
#37432 opened Apr 10, 2025 by winglian • Draft
5 tasks
1
fix document masking for chunked attention
#37429 opened Apr 10, 2025 by winglian Loading…
5 tasks
10
Refixing require_read_token
#37427 opened Apr 10, 2025 by MekkCyber Loading…
2
convert scale and zero to cuda when using HQQ backend
#37425 opened Apr 10, 2025 by phymhan Loading…
2 of 5 tasks
2
[bug] fsdp training save checkpoint twice
#37417 opened Apr 10, 2025 by cyr0930 • Draft
5 tasks
2
guard on model.eval when using torch.compile + FSDP2
#37413 opened Apr 10, 2025 by winglian Loading…
5 tasks
5
Add fuyu Fast Image Processor
#37410 opened Apr 10, 2025 by Aravind-11 • Draft
4 tasks done
1
1
chore: standardize DeBERTa model card
#37409 opened Apr 10, 2025 by Shoumik-Gandre • Draft
5 tasks done
1
[Flex Attn] Fix torch 2.5.1 incompatibilities
#37406 opened Apr 9, 2025 by vasqu Loading…
1 of 5 tasks
4
[flax/mistral] support sliding_window: null in config Flax
#37402 opened Apr 9, 2025 by yiding • Draft
1 of 5 tasks
1
Add usage example for DINOv2
#37398 opened Apr 9, 2025 by baldassarreFe Loading…
1 of 5 tasks
4
Add support for Moonlight 16B, add aux loss for Deepseek v3 model finetuning. New model
#37397 opened Apr 9, 2025 by Kimiko-AI • Draft
2 tasks done
5
Next batch of models with removed return_dict
#37396 opened Apr 9, 2025 by qubvel • Draft
5 tasks
1
[Cache] Support compilable cache reuse with smaller batch sizes
#37394 opened Apr 9, 2025 by gante Loading…
9
ProTip! Adding no:label will show everything without a label.