-
Notifications
You must be signed in to change notification settings - Fork 32.9k
Pull requests: huggingface/transformers
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
chore(typing): added modeling_utils to ty
#45425
opened Apr 14, 2026 by
tarekziade
Collaborator
Loading…
Drop
content=None from messages in apply_chat_template
#45422
opened Apr 14, 2026 by
qgallouedec
Member
Loading…
Improve nested
base_model_prefix handling in weight conversion and loading
#45421
opened Apr 13, 2026 by
yonigozlan
Member
Loading…
[serve] Forward
tool_calls/tool_call_id in processor inputs
#45418
opened Apr 13, 2026 by
qgallouedec
Member
Loading…
Adds type checking to
src/transformers/*py
#45415
opened Apr 13, 2026 by
tarekziade
Collaborator
Loading…
Fix the response schema for the gemma4 converter
#45411
opened Apr 13, 2026 by
Rocketknight1
Member
Loading…
from_pretrained orchestration + distributed save/load
#45409
opened Apr 13, 2026 by
3outeille
Member
Loading…
4 tasks
MoE expert parallelism + sequence parallelism
#45408
opened Apr 13, 2026 by
3outeille
Member
Loading…
3 tasks
avoid wrap 4bit-quantized model into DP
#45407
opened Apr 13, 2026 by
kaixuanliu
Contributor
Loading…
Fix ZeRO-3 from_pretrained: load registered buffers in _load_state_dict_into_zero3_model
#45402
opened Apr 13, 2026 by
saslifat-gif
Loading…
Add support for Voxtral-4B-TTS-2603 to transformers
Audio
New model
#45401
opened Apr 13, 2026 by
sachinkumarsingh092
•
Draft
4 of 6 tasks
Fix Qwen2.5VL temporal grid positions
for patch
Tag issues / labels that should be included in the next patch
#45400
opened Apr 13, 2026 by
zucchini-nlp
Member
Loading…
Add example for iterative chatting with MLLMs
#45398
opened Apr 13, 2026 by
zucchini-nlp
Member
Loading…
Extract dynamic vision/audio tensors into standalone pure functions
#45396
opened Apr 13, 2026 by
IlyasMoutawwakil
Member
Loading…
1 of 6 tasks
Require input_ids for repetition penalty
#45389
opened Apr 13, 2026 by
ruben-aghayan
Loading…
3 of 6 tasks
Make Gemma4ClippableLinear inherit from nn.Linear for PEFT/LoRA compatibility
#45388
opened Apr 12, 2026 by
albertorkive
Loading…
[GGUF] Reduce peak RAM usage by casting dequantized tensors early during load
#45386
opened Apr 12, 2026 by
UsamaKenway
Loading…
3 of 6 tasks
Previous Next
ProTip!
Mix and match filters to narrow down what you’re looking for.