-
Notifications
You must be signed in to change notification settings - Fork 27.5k
Issues: huggingface/transformers
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Author
Label
Projects
Milestones
Assignee
Sort
Issues list
tokenizer.decode() and tokenizer.convert_ids_to_tokens() return different results
bug
#35641
opened Jan 12, 2025 by
thangld201
4 tasks
Expected
tensors and new_tensors to have the same type but found <class ‘tuple’> and <class ‘torch.Tensor’>
bug
#35640
opened Jan 12, 2025 by
Bruce-Azar-Wayne
4 tasks
set_initialized_submodules too slow when loading big model like DeepSeekV3
bug
#35635
opened Jan 12, 2025 by
hongchuan666
4 tasks
ValueError: MllamaForConditionalGeneration does not support Flash Attention 2.0 yet
bug
#35634
opened Jan 12, 2025 by
yxchng
4 tasks
[i18n-<languageCode>] Translating docs to <languageName>
WIP
Label your PR/Issue with WIP for some long outstanding Issues/PRs that are work in progress
#35630
opened Jan 11, 2025 by
lawchingman
10 tasks
Dose
num_logits_to_keep in model.generate() really work?
bug
#35629
opened Jan 11, 2025 by
bwnjnOEI
2 of 4 tasks
static cache with mixtral will cause CUDA error: device-side assert triggered
bug
#35626
opened Jan 11, 2025 by
zyxiyy
1 of 4 tasks
The Phi model does not have lm_head bias after upgraded to v4.48.0
bug
#35625
opened Jan 11, 2025 by
yuxianq
2 of 4 tasks
Segmentation fault: address not mapped to object at address 0x100000007
bug
#35624
opened Jan 11, 2025 by
mrinaldi97
4 tasks done
Unsupported: hasattr SkipFunctionVariable when i compile the mixtral model with muti-gpus
bug
#35623
opened Jan 11, 2025 by
zyxiyy
4 tasks
The argument "dim" is gone from LlamaRotaryEmbedding initializer. Intentional?
bug
#35621
opened Jan 11, 2025 by
jeffhataws
4 tasks
from_pretrained fails to save weights.py and layers.py into cache, therefore fails to find them in cache
bug
#35619
opened Jan 11, 2025 by
openyk
4 tasks
Help Understanding Beam Search Scores in Hugging Face (LLaMA + LoRA)
bug
#35618
opened Jan 10, 2025 by
pratcooper
2 of 4 tasks
Better handeling of hardcoded component in PretrainedModel.from_pretrained.
bug
#35617
opened Jan 10, 2025 by
princethewinner
1 of 4 tasks
Trainer: TensorBoardCallback not working for "on_save" and "on_save_end" events
bug
#35612
opened Jan 10, 2025 by
vecorro
2 of 4 tasks
Trainer sets
state.best_model_checkpoint even when it doesn't save there; leads to training crash
bug
#35609
opened Jan 10, 2025 by
tomaarsen
2 of 4 tasks
Prompt_ids feature causing repetitions and hallucinations
bug
#35603
opened Jan 10, 2025 by
vchagari
4 tasks
LlavaNextVideoProcessor -> TypeError: LlavaNextVideoProcessor.__call__() got an unexpected keyword argument 'legacy' (I have the fix)
bug
Core: Pipeline
Internals of the library; Pipeline.
VLM
#35602
opened Jan 10, 2025 by
inf3rnus
2 tasks
weird criterion to decide if needed to adjust the padding size
#35599
opened Jan 9, 2025 by
hyusterr
Inconsistent saving of tokenizer with custom code from HF hub vs. local directory
bug
#35597
opened Jan 9, 2025 by
mtm-cai
2 of 4 tasks
flash_attention_2 2.7.2.post1 seems to crash when using
torch.compile and DataCollatorWithFlattening
bug
#35588
opened Jan 9, 2025 by
avishaiElmakies
4 tasks
Previous Next
ProTip!
Follow long discussions with comments:>50.