Skip to content

Issues: microsoft/DeepSpeed

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

798 Open 1,490 Closed
798 Open 1,490 Closed
Author
Filter by author
Label
Filter by label
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Milestones
Filter by milestone
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

[BUG] DeepSpeed Hybrid Engine Does not Work for Mistral-7B bug Something isn't working deepspeed-chat Related to DeepSpeed-Chat
#4954 opened Jan 14, 2024 by liziniu
[Queston] Deepspeed stage 3 OOM on 7B-model with A100-80GB memory card bug Something isn't working training
#4953 opened Jan 13, 2024 by candygocandy
[BUG] Step 3 with ZeRO=3 see error: RuntimeError: CUDA error: an illegal memory access was encountered bug Something isn't working deepspeed-chat Related to DeepSpeed-Chat
#4945 opened Jan 12, 2024 by N33MO
1
[BUG] Setting Finetune=True causes checkpoint loading to not work correctly bug Something isn't working training
#4944 opened Jan 12, 2024 by exnx
@loadams
1
AttributeError: type object 'Init' has no attribute 'quantizer_module' bug Something isn't working training
#4943 opened Jan 12, 2024 by ZetangForward
Grad parameters are None with bigger autograd graphs bug Something isn't working training
#4941 opened Jan 11, 2024 by miguelscarv
[BUG] Running DDP with transformers integrated deepspeed get a deadlock (long time no response) when training model. bug Something isn't working training
#4933 opened Jan 11, 2024 by PommesPeter
Enable deepspeed.zero.Init causes very strange spikes in PPO policy_loss
#4932 opened Jan 11, 2024 by wuxibin89
2
[BUG]deepspeed zero3 gets error in dist.get_rank() in multiple node and multiple gpu bug Something isn't working compression
#4931 opened Jan 11, 2024 by janenie
1
[REQUEST] Set num_local_io_workers bug Something isn't working
#4925 opened Jan 10, 2024 by Yangr116
[BUG] gan stage2 training TypeError: 'NoneType' object is not subscriptable bug Something isn't working training
#4923 opened Jan 9, 2024 by cat6523
[BUG] DeepSpeed Zero3 Inference behavior error when model.train() mode bug Something isn't working inference
#4922 opened Jan 9, 2024 by liu-zichen
nv-torch-nightly-v100 CI test failure ci-failure
#4921 opened Jan 9, 2024 by github-actions bot
@loadams
Get full gradients in Deepspeed 0.6.0
#4917 opened Jan 8, 2024 by leenachennuru
Cannot install deepspeed 0.12.6, fail to produce metadata. bug Something isn't working build Improvements to the build and testing systems.
#4914 opened Jan 8, 2024 by simonou99
@loadams
1
[BUG] DeepSpeed Zero Inference (stage 3) Stuck When One Process Doesn't Execute model.generate() bug Something isn't working inference
#4910 opened Jan 6, 2024 by samuel21119
2
Killing subprocess when saving checkpoints during training with zero2. bug Something isn't working training
#4909 opened Jan 6, 2024 by IvoryTower800
[BUG] my autocast is not working bug Something isn't working training
#4908 opened Jan 6, 2024 by YooSungHyun
2
[BUG] Training hang on when use zero2+cpu offload bug Something isn't working training
#4905 opened Jan 6, 2024 by boundles
BF16 training benchmarks.
#4904 opened Jan 5, 2024 by ethansmith2000
1
[BUG] ZERO++ | AssertionError: ZeRO parameter intra parallel group is already initialized bug Something isn't working training
#4901 opened Jan 5, 2024 by dhkim0225
[TASK] Seperate AutoTP workflow enhancement New feature or request
#4894 opened Jan 4, 2024 by delock
@mrwyattii
5
[BUG] Does DeepSpeed init_inference allow two engines? bug Something isn't working inference
#4893 opened Jan 4, 2024 by MinghaoYan
@mrwyattii
1
[BUG] Error: Attempting to get amgpu ISA Details 'NoneType' object has no attribute 'group' bug Something isn't working build Improvements to the build and testing systems.
#4891 opened Jan 3, 2024 by unavailableun
@loadams
3
[BUG] [ERROR] [launch.py:321:sigkill_handler [xxx] exits with return code = -9 bug Something isn't working training
#4890 opened Jan 3, 2024 by xinbingzhe
ProTip! no:milestone will show everything without a milestone.