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

814 Open 6,325 Closed
814 Open 6,325 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 interrupting a run with Ctrl+C, sometimes the WandbLogger does not upload a checkpoint artifact bug Something isn't working needs triage Waiting to be triaged by maintainers ver: 2.4.x
#20425 opened Nov 16, 2024 by edmcman
2
Why only one GPU is getting used in the kaggle kernel bug Something isn't working needs triage Waiting to be triaged by maintainers ver: 2.4.x
#20424 opened Nov 16, 2024 by KeesariVigneshwarReddy
Weird error while training a model with tabular data!!!! Some problem related self.log_dict bug Something isn't working needs triage Waiting to be triaged by maintainers ver: 2.4.x
#20423 opened Nov 16, 2024 by KeesariVigneshwarReddy
PyTorch Lightning uses deprecated torch_xla function, causing compatibility issues with latest torch_xla versions bug Something isn't working needs triage Waiting to be triaged by maintainers ver: 2.4.x
#20419 opened Nov 13, 2024 by aahila-aws
Log default metrics feature Is an improvement or enhancement needs triage Waiting to be triaged by maintainers
#20418 opened Nov 13, 2024 by ierezell
seed_everything(..., workers=True) causes the Dataloader to apply exactly the same augmentations each epoch if they sample values from torch.distributions bug Something isn't working needs triage Waiting to be triaged by maintainers ver: 2.4.x
#20412 opened Nov 12, 2024 by nan-dre
1
update dataset at "on_train_epoch_start", but "training_step" still get old data bug Something isn't working loops Related to the Loop API
#20407 opened Nov 8, 2024 by Yak1m4Sg
1
FSDP full state dict mangles fsspec path bug Something isn't working needs triage Waiting to be triaged by maintainers ver: 2.4.x
#20406 opened Nov 8, 2024 by oceanusxiv
How to deal with uneven inputs in DDP with sharded data without hanging bug Something isn't working needs triage Waiting to be triaged by maintainers ver: 2.4.x
#20404 opened Nov 7, 2024 by ssharpe42
Proposal(CLI): after_instantiate_classes hook design Includes a design discussion feature Is an improvement or enhancement lightningcli pl.cli.LightningCLI
#20400 opened Nov 6, 2024 by AlessandroW
1
PytorchStreamReader failed reading zip archive: not a ZIP archive bug Something isn't working needs triage Waiting to be triaged by maintainers ver: 2.4.x
#20398 opened Nov 6, 2024 by Crazy-LittleBoy
2
put the monitor metric into default filename for ModelCheckpoint feature Is an improvement or enhancement needs triage Waiting to be triaged by maintainers
#20397 opened Nov 5, 2024 by VDFaller
Light / dark mode for documentation docs Documentation related needs triage Waiting to be triaged by maintainers
#20396 opened Nov 5, 2024 by nbrosse
3
Gradient checkpointing and ddp do not work together bug Something isn't working needs triage Waiting to be triaged by maintainers ver: 2.4.x
#20395 opened Nov 4, 2024 by rubenweitzman
Error if SLURM_NTASKS != SLURM_NTASKS_PER_NODE bug Something isn't working needs triage Waiting to be triaged by maintainers ver: 2.4.x
#20391 opened Nov 4, 2024 by guarin
Major performance degradation when multiple metrics/losses bug Something isn't working ver: 2.4.x
#20388 opened Nov 3, 2024 by EtayLivne
2
FSDP with HYBRID_SHARD loss doesn't improve with more nodes bug Something isn't working needs triage Waiting to be triaged by maintainers ver: 2.4.x
#20385 opened Nov 2, 2024 by zaptrem
Custom TQDMProgressBar changes not reflected bug Something isn't working needs triage Waiting to be triaged by maintainers ver: 2.4.x
#20384 opened Nov 1, 2024 by oseymour
Optimize fit_loop() to reduce train_dataloader()'s memory footprint feature Is an improvement or enhancement repro needed The issue is missing a reproducible example
#20382 opened Nov 1, 2024 by guillaume-rochette-oxb
2
Fabric and FFCV? repro needed The issue is missing a reproducible example
#20380 opened Nov 1, 2024 by richardrl
1
Deepspeed ZERO MiCS support feature Is an improvement or enhancement needs triage Waiting to be triaged by maintainers
#20378 opened Oct 31, 2024 by hehepig4
FSDP checkpoint loading fails bug Something isn't working needs triage Waiting to be triaged by maintainers ver: 2.4.x
#20373 opened Oct 29, 2024 by Nilabhra
metrics csv in ddp mode bug Something isn't working needs triage Waiting to be triaged by maintainers ver: 2.2.x
#20371 opened Oct 29, 2024 by ruyanyinian
FutureWarning: torch.cuda.amp.custom_bwd(args...) is deprecated. Please use torch.amp.custom_bwd(args..., device_type='cuda') instead. bug Something isn't working needs triage Waiting to be triaged by maintainers ver: 2.4.x
#20370 opened Oct 28, 2024 by loretoparisi
Wandb 1.x step handling feature Is an improvement or enhancement needs triage Waiting to be triaged by maintainers
#20368 opened Oct 28, 2024 by edmcman
ProTip! Adding no:label will show everything without a label.