Repository Issues
Lightning-AI/pytorch-lightning
Pretrain, finetune and deploy AI models on multiple GPUs, TPUs with zero code changes.
Issues
Open
BatchSizeFinder throws KeyError: 'limit_eval_batches'
bugduplicatehelp wantedrepro neededtunerver: 2.1.x
No assignee yetHas a beginner-friendly labelContributing guide available
3 comments1 reaction0 assignees
Open
registered buffers' dtype is overridden after __init__
fabricfeaturehelp wantedplprecision: halfver: 2.1.x
Why recommendedNo assignee yet · Has a beginner-friendly label
No assignee yetHas a beginner-friendly labelContributing guide available
3 comments0 reactions0 assignees
Open
Support for returning LRSchedulerConfig on LightningModule.configure_optimizers
featurehelp wantedlightningmodulelr scheduler
Why recommendedNo assignee yet · Has a beginner-friendly label
No assignee yetHas a beginner-friendly labelContributing guide available
1 comment0 reactions0 assignees
Open
Support saving and loading remote paths with FSDP
checkpointingfeaturehelp wantedstrategy: fsdp
Why recommendedNo assignee yet · Has a beginner-friendly label
No assignee yetHas a beginner-friendly labelContributing guide available
5 comments1 reaction0 assignees
Open
Deepspeed activation Partitioning
docshelp wantedstrategy: deepspeed
Why recommendedNo assignee yet · Has a beginner-friendly label
No assignee yetHas a beginner-friendly labelContributing guide available
2 comments0 reactions0 assignees
Open
Adopt `typing_extensions.Override`
featuregood first issuehelp wanted
Why recommendedHas a beginner-friendly label · Contributing guide available
Has a beginner-friendly labelContributing guide available
25 comments0 reactions1 assignee
Open
Provide a non-expert mode for the DeviceStatsMonitor
callback: device statsfeaturehelp wanted
Why recommendedNo assignee yet · No comments yet
No assignee yetNo comments yetHas a beginner-friendly labelContributing guide available
0 comments15 reactions0 assignees
Open
Why does running Lightning on SLURM with python perform worse than with srun?
environment: slurmhelp wantedquestionver: 2.0.x
Why recommendedNo assignee yet · Has a beginner-friendly label
No assignee yetHas a beginner-friendly labelContributing guide available
21 comments2 reactions0 assignees
Open
Spurious validation step when restarting with a checkpoint when `max_steps` is set in the trainer
bughelp wantedloopsver: 2.0.x
Why recommendedNo assignee yet · Has a beginner-friendly label
No assignee yetHas a beginner-friendly labelContributing guide available
1 comment3 reactions0 assignees
Open
Investigate Resident Memory Increase during Inference
bughelp wantedperformancever: 2.0.x
Why recommendedNo assignee yet · Has a beginner-friendly label
No assignee yetHas a beginner-friendly labelContributing guide available
9 comments8 reactions0 assignees
Open
W&B `dir` flag doesn't work.
bughelp wantedlogger: wandbver: 2.0.x
Why recommendedNo assignee yet · Has a beginner-friendly label
No assignee yetHas a beginner-friendly labelContributing guide available
4 comments3 reactions0 assignees
Open
`CombinedLoader` takes a long time when `num_workers > 0`
bughelp wantedperformancerepro neededver: 2.0.x
Why recommendedNo assignee yet · Has a beginner-friendly label
No assignee yetHas a beginner-friendly labelContributing guide available
1 comment0 reactions0 assignees
Open
TPU profiler
docshelp wantedprofiler
Why recommendedNo assignee yet · Has a beginner-friendly label
No assignee yetHas a beginner-friendly labelContributing guide available
2 comments0 reactions0 assignees
Open
Can't run the pytorch lightning program packaged with pyinstaller.
3rd partybughelp wantedrepro neededver: 1.9.x
Why recommendedNo assignee yet · Has a beginner-friendly label
No assignee yetHas a beginner-friendly labelContributing guide available
3 comments4 reactions0 assignees
Open
MLflow logger to log config.yaml as artifact
featurehelp wantedlogger: mlflow
Why recommendedNo assignee yet · Has a beginner-friendly label
No assignee yetHas a beginner-friendly labelContributing guide available
3 comments1 reaction0 assignees
Open
Lightning CLI should use fail_untyped=True
3rd partybreaking changehelp wantedlightningclirefactor
Why recommendedHas a beginner-friendly label · Contributing guide available
Has a beginner-friendly labelContributing guide available
4 comments1 reaction1 assignee
Open
Resuming from checkpoint, mid epoch gives a very distorted time estimate
bughelp wantedpriority: 1progress bar: tqdmver: 2.0.x
Why recommendedNo assignee yet · Has a beginner-friendly label
No assignee yetHas a beginner-friendly labelContributing guide available
2 comments1 reaction0 assignees
Open
TPU v3-8 deadlocks when using datasets larger than 2^15 on 8 devices
accelerator: tpubughelp wantedstrategy: xlaver: 2.0.x
Why recommendedNo assignee yet · Has a beginner-friendly label
No assignee yetHas a beginner-friendly labelContributing guide available
8 comments0 reactions0 assignees
Open
Incorrect batch progress saved in checkpoint at every_n_train_steps
bughelp wantedloopsrepro neededver: 1.9.xver: 2.1.x
Why recommendedNo assignee yet · Has a beginner-friendly label
No assignee yetHas a beginner-friendly labelContributing guide available
6 comments6 reactions0 assignees
Open
Using `pyg` sampler causes lightning to skip validation step
3rd partybughelp wantedver: 2.1.x
Why recommendedNo assignee yet · No comments yet
No assignee yetNo comments yetHas a beginner-friendly labelContributing guide available
0 comments1 reaction0 assignees