issues
search
huggingface
/
transformers
🤗 Transformers: State-of-the-art Machine Learning for Pytorch, TensorFlow, and JAX.
https://huggingface.co/transformers
Apache License 2.0
132.7k
stars
26.44k
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
RuntimeError: expand(torch.FloatTensor{ ... }, size = [...]) the number of sizes provided (4) must be greater or equal to the number of dimensions in the tensor (5)
#33740
lbertge
opened
1 day ago
3
Add Flash Attention Support to Mllama (Llama 3.2)
#33739
shermansiu
closed
1 day ago
2
Fix MoE tensor reshape
#33738
sukjunhwang
opened
2 days ago
1
Fix modular model converter unable to generate Processor classes
#33737
tonywu71
closed
2 days ago
3
Add ColPali to 🤗 transformers
#33736
tonywu71
opened
2 days ago
0
[`clean_up_tokenization_spaces`] Pl bart was failing, updating
#33735
ArthurZucker
closed
1 day ago
1
fix: add docstring for `image_size` in Convnextv2 config
#33734
lucianosrp
closed
2 days ago
1
Trainer doesn't save evaluation metrics.
#33733
filbeofITK
opened
2 days ago
1
[whisper] added dropping of attention weights after DTW calculations related to word timestamps if these weights are not requested in the output
#33732
jacekc3
opened
2 days ago
1
Fix data_seed unused
#33731
MekkCyber
opened
2 days ago
1
Make audio classification pipeline spec-compliant and add test
#33730
Rocketknight1
closed
1 day ago
4
Add GLM4 model
#33729
Cyrilvallez
opened
2 days ago
1
Will Trainer.predict() return data in the same order as the original dataset during multi-machine and multi-gpus inference?
#33728
deepwebney
opened
2 days ago
1
Issue with transformers 4.45.0 and torchao 0.1 cannot import name 'quantize_' from 'torchao.quantization'
#33727
EnragedAntelope
closed
2 days ago
1
Fix docs and docstrings Omdet-Turbo
#33726
yonigozlan
closed
2 days ago
1
[PEFT] Support low_cpu_mem_usage option for PEFT loading adapters
#33725
BenjaminBossan
opened
2 days ago
2
Add sdpa for DistilBert
#33724
OmarManzoor
opened
2 days ago
1
Added Ukrainian translations
#33723
Drugak
opened
2 days ago
0
Validate the eval dataset in advance.
#33722
jackyjinjing
closed
1 day ago
0
validation of the eval dataset should be done in advance
#33721
jackyjinjing
opened
2 days ago
0
[i18n-ZH] Sync and Localize Latest English Readme to Simplified Chinese Readme
#33720
vortezwohl
opened
2 days ago
0
Misleading warning
#33719
krisztian-gajdar
opened
2 days ago
2
Generate: `can_generate()` recursive check
#33718
gante
closed
2 days ago
2
Trainer class causes massive memory leak when using mps
#33717
JamesBowerXanda
opened
2 days ago
18
add sdpa and flash_attention2 support to speech2text
#33716
avishaiElmakies
opened
2 days ago
1
[`MllamaProcessor`] Update errors and API with multiple image
#33715
ArthurZucker
closed
2 days ago
1
Stable dropout show drop prob in model print
#33714
fkrasnov2
opened
2 days ago
2
Doc and config mismatch for DeBERTa
#33713
fkrasnov2
closed
1 day ago
0
Add the LARS optimizers for training large scale CNN model with larger batch size
#33712
dame-cell
closed
1 day ago
4
Add support for custom inputs and batched inputs in ProcessorTesterMixin
#33711
yonigozlan
opened
2 days ago
2
Add support for Molmo
#33710
fakerybakery
opened
2 days ago
2
Gemma is ExecuTorch compatible
#33709
guangy10
opened
2 days ago
0
Export-to-ExecuTorch via transformers.js integration
#33708
guangy10
opened
2 days ago
2
Generate using exported model and enable gemma2-2b in ExecuTorch
#33707
guangy10
opened
2 days ago
11
`processing_mllama.py` has a bug?
#33706
Neo9061
closed
2 days ago
1
Add index selection for `output_hidden_states`
#33705
hlky
opened
3 days ago
1
Update Albumentations Versions
#33704
vasqu
closed
1 day ago
1
Add MLLama
#33703
ArthurZucker
closed
3 days ago
1
fix: use correct var names for check_tokenizers script
#33702
niqodea
closed
2 days ago
2
🌐 [i18n-KO] Translated `backbones.md` to Korean
#33701
jun048098
opened
3 days ago
0
add verification of cache position
#33700
ManuelFay
closed
2 days ago
3
Fix paligemma `eager` vs `sdpa` + image transforms test fixup
#33699
molbap
opened
3 days ago
3
Refactor `output_hidden_states` to allow index selection
#33698
hlky
opened
3 days ago
3
Different results obtained using pipeline (worse) vs. model.generate under the same decoding strategy
#33697
kirk86
opened
3 days ago
1
Refactor image features selection in LlaVa
#33696
kenza-bouzid
opened
3 days ago
3
Refactor image features selection in LlaVa
#33695
kenza-bouzid
opened
3 days ago
1
Update modeling_longformer.py so it can be converted to onnx format using Pytorch.
#33694
xiaowuhu
opened
3 days ago
0
7% of training time with `Trainer` is spent on a single line: `and (torch.isnan(tr_loss_step) or torch.isinf(tr_loss_step))`
#33693
umarbutler
opened
3 days ago
2
[WIP] Sink cache: fix implementation to shift key states
#33692
why-in-Shanghaitech
opened
3 days ago
3
The Sink Cache looks wired
#33691
why-in-Shanghaitech
opened
3 days ago
1
Previous
Next