Skip to content

Issues: huggingface/accelerate

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

101 Open 1,608 Closed
101 Open 1,608 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

When saving checkpoint in multi-node training, using Zero3 optimization is model.safetensors file full model ?
#3381 opened Feb 6, 2025 by KeshavSingh29
2 of 4 tasks
DDP hanging and timing out on H100s, not other hardware
#3380 opened Feb 6, 2025 by EvanKomp
1
estimate-memory gives ~64GB for meta-llama/Llama-3.1-70B-Instruct instead of anticipated ~140GB
#3379 opened Feb 5, 2025 by dvrogozh
AMD Suggestion, TunableOP tuning script
#3377 opened Feb 3, 2025 by Charmandrigo
Accelerate launch will hang when the training loop data loader have a file not found error (even though this is catched)
#3375 opened Feb 2, 2025 by lucasjinreal
1
Initialize model with empty weight causes OOM with offloading to disk
#3374 opened Feb 1, 2025 by Aiden-Frost
2 of 4 tasks
loading the prodigy optimizer does not move custom parameters to the accelerator
#3372 opened Jan 29, 2025 by bghira
4 tasks
Gettin gradient of loss during inference
#3371 opened Jan 28, 2025 by LalchandPandia
9
Training hangs indefinitely on first forward pass when using TPU v3-8 in Kaggle
#3370 opened Jan 27, 2025 by WpythonW
2 of 4 tasks
3
Gradient accumulation with deepSpeed has issue if not set during configuration
#3369 opened Jan 27, 2025 by khalil-Hennara
2 of 4 tasks
1
DataLoaderShard wrongly yields None instead of StopIteration when its dataloader returns StopIteration immediately
#3367 opened Jan 25, 2025 by Aleko2286
2 of 4 tasks
1
"@verify_operation" lead to pretrain of multi-nodes hang
#3364 opened Jan 24, 2025 by sankexin
2 of 4 tasks
tests/test_cli.py::ModelEstimatorTester::test_no_split_modules fails after 84a67891 in Transformers
#3362 opened Jan 23, 2025 by dvrogozh
3
FP8 training can't works with deepspeed
#3360 opened Jan 23, 2025 by XiaobingSuper
2 of 4 tasks
1
1
Google Colab TPU notebook_launcher doesn't work
#3358 opened Jan 21, 2025 by matinmoezzi
2 of 4 tasks
1
how to config accelerate on 2 mac machines
#3356 opened Jan 20, 2025 by hsoftxl
4
Questions on large model inference / finetuning
#3353 opened Jan 18, 2025 by kalyani7195
3
Stuck forever in accelerator.backward without any logs 2
#3347 opened Jan 16, 2025 by XueruiSu
2
Maybe a conflict between accelerate and transformers CLIPVisionModel
#3339 opened Jan 13, 2025 by striveAgain
2 of 4 tasks
[Feature Request] include a DeepSpeed multi-node config slurm example contributions-welcome deepspeed DS related issues/PRs
#3338 opened Jan 13, 2025 by sayakpaul
1
Accelerate with DeepSpeed not works
#3337 opened Jan 12, 2025 by khalil-Hennara
2 of 4 tasks
7
include MIGraphX dynamo backend on accelerate config
#3335 opened Jan 10, 2025 by Charmandrigo
Optimal way to train on TPU POD
#3325 opened Jan 6, 2025 by DuyguA
1
support customized dataLoaders?
#3322 opened Jan 5, 2025 by ifiaposto
1
How to save self-defined model with deepspeed zero 3?
#3320 opened Jan 2, 2025 by amoyplane
2 of 4 tasks
1
ProTip! Add no:assignee to see everything that’s not assigned.