-
Notifications
You must be signed in to change notification settings - Fork 28k
Pull requests: huggingface/transformers
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
Add missing atol to torch.testing.assert_close where rtol is specified
#36234
opened Feb 17, 2025 by
ivarflakstad
Loading…
Prevent Reinitialization of Resized LM Head When
tie_word_embeddings
is False #35141
#36221
opened Feb 16, 2025 by
sambhavnoobcoder
Loading…
fix: prevent second save in the end of training if last step was saved already
#36219
opened Feb 16, 2025 by
NosimusAI
Loading…
2 of 5 tasks
Improvements in attention_forward functions
#36218
opened Feb 15, 2025 by
mseeger
Loading…
3 of 5 tasks
[WIP] Add a dedicated tokenizer for byte level transformers
#36216
opened Feb 15, 2025 by
apehex
Loading…
Change Qwen2_VL image processors to have init and call accept the same kwargs
#36207
opened Feb 14, 2025 by
yonigozlan
Loading…
Append best model checkpoint with active adapter when not default
#36201
opened Feb 14, 2025 by
Thomas26948
Loading…
1 of 5 tasks
Fixed dynamic module import when there is more than one dit in class …
#36198
opened Feb 14, 2025 by
ExtReMLapin
Loading…
(ugly) Use
parallelism=4
for check_repository_consistency
#36197
opened Feb 14, 2025 by
ydshieh
Loading…
Qwen2VL fix cos,sin dtypes to float when used with deepspeed
#36188
opened Feb 14, 2025 by
ArdalanM
Loading…
5 tasks
Remove differences between init and preprocess kwargs for fast image processors
#36186
opened Feb 13, 2025 by
yonigozlan
Loading…
Add Got-OCR 2 Fast image processor and refactor slow one
#36185
opened Feb 13, 2025 by
yonigozlan
Loading…
Try working around the processor registration bugs
#36184
opened Feb 13, 2025 by
Rocketknight1
•
Draft
[CI] Check test if the
GenerationTesterMixin
inheritance is correct 🐛 🔫
#36180
opened Feb 13, 2025 by
gante
Loading…
Previous Next
ProTip!
Follow long discussions with comments:>50.