Skip to content

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.

Sign up for GitHub

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

998 Open 15,351 Closed
998 Open 15,351 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
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

Maybe the way SequenceClassification Model calculates the last non-pad token is not reasonable. bug
#35352 opened Dec 20, 2024 by liangxuZhang
4 tasks
Any plans to add AIMv2 in the model? New model
#35351 opened Dec 20, 2024 by nahidalam
2 tasks done
SinkCache (StreamLLM) implemented over Post-RoPE Key cache might result in confused position for inference bug
#35350 opened Dec 19, 2024 by wangguangtao0722
4 tasks
A warning message showing that MultiScaleDeformableAttention.so is not found in /root/.cache/torch_extensions if ninja is installed with transformers bug
#35349 opened Dec 19, 2024 by cainmagi
1 of 4 tasks
[Mamba2] Varlen implementation Feature request Request for a new feature
#35346 opened Dec 19, 2024 by vasqu
1
Llama model, torch.compile output for custom device does not match with eager/cpu when generation_config.use_cache set to True bug
#35343 opened Dec 19, 2024 by vpandya-quic
4 tasks
Option to Disable Model Caching When Using "pipeline" Feature request Request for a new feature
#35337 opened Dec 19, 2024 by FadiAmon
1
Default arguments in DebertaConfig disable relative attention, contrary to the docs and deberta-base bug
#35335 opened Dec 19, 2024 by bauwenst
4 tasks
4
AllAboardBertweetModel New model
#35333 opened Dec 19, 2024 by CharC0de
2 tasks done
DeBERTa's DisentangledSelfAttention hardcodes float dtype, which causes bfloat16 overflow error bug
#35332 opened Dec 19, 2024 by bauwenst
2 of 4 tasks
1
MPI environment variables are not set. bug
#35331 opened Dec 18, 2024 by fabiogeraci
2 of 4 tasks
2
tokenizer decode decode with timestamp fails for extended vocabulary bug
#35330 opened Dec 18, 2024 by bnestor
2 of 4 tasks
1
InternVL is ExecuTorch Compatible ExecuTorch Feature request Request for a new feature
#35327 opened Dec 18, 2024 by guangy10
unable to convert llama 3.3 weights to hf.py bug
#35326 opened Dec 18, 2024 by AshishMulupuri
1 of 4 tasks
Deepseek v2 New model
#35317 opened Dec 18, 2024 by VladOS95-cyber
2 tasks done
2
train_new_from_iterator() does not work when pre_tokenizer is null bug
#35315 opened Dec 18, 2024 by cecheta
1 of 4 tasks
2
Unclear what happens when using torchrun, multi-gpu and trainer arguments. bug
#35311 opened Dec 17, 2024 by davies-w
2 of 4 tasks
Multi-GPU training crashes with IterableDataset and different length input (e.g. Next token prediction) bug
#35308 opened Dec 17, 2024 by avishaiElmakies
2 of 4 tasks
1
[Question] Why doesn't trainer.state.epoch fall round after training? trainer
#35298 opened Dec 16, 2024 by qgallouedec
Custom 4D tensor caused shape mismatch error bug
#35290 opened Dec 16, 2024 by fingertap
1 of 4 tasks
version 4.47.0 provides different generation results when using quantized awq model bug
#35286 opened Dec 16, 2024 by xin3he
2 of 4 tasks
Request to add D-FINE contributions-welcome Good Second Issue Issues that are more difficult to do than "Good First" issues - give it a try if you want! New model Vision
#35283 opened Dec 15, 2024 by brockt96
1 of 2 tasks
13
Qwen2vl support for GGUF Feature request Request for a new feature
#35282 opened Dec 15, 2024 by cheald
2
Vision models don't work for non-square object bug Vision
#35280 opened Dec 15, 2024 by liujilei156231
2 of 4 tasks
1
KeyError: 'intern_vit_6b' bug Multimodal
#35279 opened Dec 15, 2024 by Sweewangyu
1 of 4 tasks
1
ProTip! Find all open issues with in progress development work with linked:pr.