Skip to content

Issues: Lightning-AI/pytorch-lightning

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

Clear current search query, filters, and sorts
34 Open 116 Closed
34 Open 116 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

enable loading universal checkpointing checkpoint in DeepSpeedStrategy feature Is an improvement or enhancement help wanted Open to be worked on strategy: deepspeed
#20065 opened Jul 9, 2024 by zhoubay
1
When calling trainer.test() train_dataloader is also validated, which makes no sense bug Something isn't working strategy: deepspeed
#19745 opened Apr 8, 2024 by asusdisciple
2
batch_sampler.batch_size is None with deepspeed and DataLoader(batch_size=None) bug Something isn't working help wanted Open to be worked on strategy: deepspeed
#19460 opened Feb 13, 2024 by olegsinavski 2.4.x
4
Add additional parameters to the DeepSpeedStrategy feature Is an improvement or enhancement strategy: deepspeed
#19278 opened Jan 12, 2024 by keunwoochoi
1
Deepspeed vs DDP bug Something isn't working strategy: deepspeed ver: 2.1.x
#19246 opened Jan 8, 2024 by jpatel-bdai
2
Deepspeed Stage 3 crashes Lightning trainer bug Something isn't working repro needed The issue is missing a reproducible example strategy: deepspeed ver: 2.1.x
#19096 opened Nov 30, 2023 by m-harmonic
2
Unable to chnage checkpoint in on_save_checkpoint with Deepspeed bug Something isn't working checkpointing Related to checkpointing strategy: deepspeed ver: 2.0.x
#18747 opened Oct 8, 2023 by xluo233 future
5
Deepspeed activation Partitioning docs Documentation related help wanted Open to be worked on strategy: deepspeed
#18732 opened Oct 6, 2023 by LogicBaron
2
Model trained with Deepspeed stage 3 shape not match when loading bug Something isn't working repro needed The issue is missing a reproducible example strategy: deepspeed ver: 2.0.x
#18648 opened Sep 26, 2023 by yinweisu
2
Device error when loading from checkpoint for testing with deepspeed bug Something isn't working strategy: deepspeed ver: 2.0.x
#18478 opened Sep 4, 2023 by dionman
1
Memory Leak when instantiating Fabric multiple times bug Something isn't working fabric lightning.fabric.Fabric performance strategy: deepspeed ver: 2.0.x
#18356 opened Aug 21, 2023 by vkakerbeck
3
CUDA OOM with DeepSpeed ZeRO Stage 3 Offload bug Something isn't working strategy: deepspeed ver: 2.1.x
#18134 opened Jul 21, 2023 by rggs
4
Any plans in adding support for DeepSpeedHybridEngine feature Is an improvement or enhancement strategy: deepspeed
#17682 opened May 23, 2023 by ShaojieJiang
--hf_deepspeed_save flag to use Hugging Face Deepspeed logic and no configure_optimizers if optimizer/scheduler defined feature Is an improvement or enhancement strategy: deepspeed
#17673 opened May 21, 2023 by jamesharrisivi
Update deepspeed activation checkpointing docs docs Documentation related help wanted Open to be worked on strategy: deepspeed
#17621 opened May 12, 2023 by avivbrokman
1
Error using lightning 2.0 when i use deepspeed and torch.compile both 3rd party Related to a 3rd-party bug Something isn't working strategy: deepspeed ver: 2.1.x
#17549 opened May 3, 2023 by yw0nam
Refactor the DeepSpeed strategy config management fabric lightning.fabric.Fabric pl Generic label for PyTorch Lightning package refactor strategy: deepspeed
#17472 opened Apr 25, 2023 by carmocca future
1
deepspeed strategy can't save checkpoint, TypeError: cannot pickle torch._C._distributed_c10d.ProcessGroup object 3rd party Related to a 3rd-party bug Something isn't working repro needed The issue is missing a reproducible example strategy: deepspeed ver: 2.0.x
#17369 opened Apr 13, 2023 by dmitrymailk
12
Deepspeed stage 3 crashing with student + teacher question Further information is requested strategy: deepspeed ver: 2.0.x
#17319 opened Apr 10, 2023 by andrasiani
1
when using huggingface pretrained model with multi-gpu, model parameters were duplicate for every gpu in ram 3rd party Related to a 3rd-party question Further information is requested strategy: deepspeed waiting on author Waiting on user action, correction, or update
#17043 opened Mar 12, 2023 by linyubupa
19
Integration with DeepSpeed and PyG: expected scalar type Float but found Half 3rd party Related to a 3rd-party bug Something isn't working strategy: deepspeed
#16793 opened Feb 17, 2023 by adm995
Examples for OPT, just like benchmark on minGPT example feature Is an improvement or enhancement strategy: deepspeed
#16671 opened Feb 7, 2023 by young-chao
trainer.save_checkpoint doesn't work after trainer.test with deepspeed strategy checkpointing Related to checkpointing strategy: deepspeed trainer: test trainer: validate
#15247 opened Oct 22, 2022 by rohitgr7 future
mixed precision with Deepspeed bug Something isn't working help wanted Open to be worked on strategy: deepspeed
#15168 opened Oct 18, 2022 by wangleiofficial
6
trainer.validate() will not load optimizer properly, different behavior from trainer.fit() strategy: deepspeed
#14993 opened Oct 4, 2022 by MattYoon
2
ProTip! Mix and match filters to narrow down what you’re looking for.