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

99 Open 1,671 Closed
99 Open 1,671 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

Link to Accelerate’s internal mechanisms doesn't seem to work
#3506 opened Apr 15, 2025 by alvaro-mazcu
1
2
prepare_pippy raises StopIteration, due to generate_device_map doesn't cover all gpus
#3500 opened Apr 11, 2025 by wangzizhao
2 of 4 tasks
1
acclerate launch failed for multi-node + multi-process on azure machine learning clusters
#3494 opened Apr 8, 2025 by ZhiliangWu
2 of 4 tasks
1
FSDP2 Unwrap model, still has dtensors as weights bug Something isn't working
#3487 opened Apr 5, 2025 by kiansierra
2 of 4 tasks
3
Minimal FSDP2 w Accelerate + SFTTrainer doesn't properly set DeviceMesh bug Something isn't working
#3486 opened Apr 5, 2025 by winglian
2 of 4 tasks
6
PartialState split_between_processes resulting in duplicated output with padding=True bug Something isn't working
#3485 opened Apr 4, 2025 by zcakzhuu
1 of 4 tasks
1
1
DeepSpeed ZeRO Stage 2/3 incompatibility with no_sync context manager
#3481 opened Apr 4, 2025 by alisafaya
2
max_memory key can not be "disk"
#3480 opened Apr 3, 2025 by Garbaz
2 of 4 tasks
5
[Discussion?] about converting outputs to fp32
#3477 opened Apr 3, 2025 by cyr0930
2
Is Fsdp2 optimization slower?
#3473 opened Apr 1, 2025 by kadirnar
4
More support for Pipeline Parallism PP in accelearte?
#3468 opened Mar 31, 2025 by dragen1860
1
Accelerate does not propagate dynamic=False to torch.compile
#3467 opened Mar 29, 2025 by Aleko2286
2 of 4 tasks
Unknown num_process when using AMAZON_SAGEMAKER
#3463 opened Mar 27, 2025 by GeraudK
2 of 4 tasks
Zero-3, model copy
#3458 opened Mar 26, 2025 by jiangxinke
4 tasks
1
Loading error with bnb 4bit config
#3455 opened Mar 22, 2025 by vyi
LOCAL_RANK environment variable issue
#3454 opened Mar 22, 2025 by chunlexiong-ai
Issue when training with MPS
#3453 opened Mar 21, 2025 by louisbrulenaudet
1 of 4 tasks
1
Add support for vllm library name in estimate-memory
#3452 opened Mar 21, 2025 by baggiponte
1 of 4 tasks
accelerate Fails with 3 GPUs but Works with 2 GPUs
#3446 opened Mar 19, 2025 by kuri54
2 of 4 tasks
1
Add Support for MiCS Shards on Deepspeed
#3442 opened Mar 13, 2025 by sam-h-bean
1
Issue with distribution strategy for sagemaker
#3440 opened Mar 13, 2025 by felix-dumont-neoxia
1
Using Deepspeed zero 3 to save Checkpoint
#3435 opened Mar 11, 2025 by JYX1216
2 of 4 tasks
1
Issue with model generations and using GRPOTrainer with device_map='auto' passed while loading model.
#3434 opened Mar 10, 2025 by debdeepsanyal
2 of 4 tasks
3
Could not load random states in Pytorch >= 2.4
#3433 opened Mar 10, 2025 by XiaoyuBIE1994
2 of 4 tasks
1
Slowdown caused by accelerate or pytorch?
#3431 opened Mar 9, 2025 by ZeshengLiu22
1
ProTip! Follow long discussions with comments:>50.