Go
Home
Pricing
FAQ
Install
Home
Pricing
FAQ
Install
Login
via GitHub
Lightning-AI/pytorch-lightning
Pull Requests
Commits
Open
Closed
Improve error message for mixed strategy namespaces
#22017 opened 2026-10-09 16:14 by
Subodh-17
fix(strategies): report the world size as `num_replicas` for DDP and FSDP
#22016 opened 2026-10-09 14:49 by
dai123x
feat(loggers): add group_metrics option to TensorBoardLogger
#22015 opened 2026-10-09 12:16 by
dai123x
LightningCLI: avoid jsonargparse v4 deprecation warnings and prepare for v5
#22014 opened 2026-10-09 07:45 by
mauvilsa
Add `from_fx()` and `to_unfx()` utilities for fx-converted models
#22013 opened 2026-10-09 04:46 by
KeepItJibreel
Reset `Timer` end timestamps when a stage restarts
#22012 opened 2026-10-08 17:23 by
aalquwayfili
ci: pin runner images to ubuntu-24.04, macos-26, windows-2025
#22010 opened 2026-10-07 18:57 by
bhimrazy
feat(fabric): add return-type overloads for Fabric.setup, setup_optimizers and setup_dataloaders
#22009 opened 2026-10-07 10:51 by
dai123x
fix: make standalone returns failure when its test runner fails
#22008 opened 2026-10-07 05:07 by
dai123x
docs: highlight logger API examples with emphasize-lines
#22007 opened 2026-10-06 17:39 by
Darshankg18
fix: set sampler epoch before creating resumed iterator
#22006 opened 2026-10-04 16:23 by
Subodh-17
fix(profiler): gracefully handle stopping unstarted actions in Advanc…
#22005 opened 2026-10-04 11:10 by
dai123x
Fix trainer.test() logging incorrect epoch when restoring from checkpoint
#22004 opened 2026-10-04 06:31 by
aryansirohi2121-coder
Preserve first-loader metrics across evaluation reloads
#22002 opened 2026-10-03 13:22 by
mukeshr06
Accept callback sequences without mutating caller inputs
#22001 opened 2026-10-03 11:48 by
mukeshr06
Use recorded-loss count in LR finder searches
#22000 opened 2026-10-03 10:17 by
mukeshr06
Resume completed epochs normally when max_steps is set
#21999 opened 2026-10-03 05:14 by
mukeshr06
Size Rich metric columns by terminal display width
#21998 opened 2026-10-03 01:10 by
mukeshr06
Honor explicit repeated PyTorch profiler schedules
#21997 opened 2026-10-02 23:54 by
mukeshr06
Preserve async checkpoint save/remove ordering
#21996 opened 2026-10-02 18:16 by
mukeshr06
Make MNIST helper tests independent of downloads
#21995 opened 2026-10-02 17:48 by
mukeshr06
Fix inflated tqdm rates after resuming mid-epoch
#21994 opened 2026-10-02 17:16 by
mukeshr06
Fix CLI config saving without a local logger directory
#21993 opened 2026-10-02 15:32 by
aswanth-07
Fix EMA updates after skipped AMP optimizer steps
#21991 opened 2026-10-01 20:02 by
anishhongekar47-dev
docs: fix typo in `trainer.rst`
#21989 opened 2026-10-01 17:33 by
adosar
fix: AMP autocast device for accelerator instances and custom registered accelerators
#21986 opened 2026-09-30 08:29 by
deependujha
Fix FitLoop leaving _restart_stage in RESTARTED_MID_EPOCH after fast-forwarding an epoch
#21985 opened 2026-09-30 04:57 by
mayuriphad
Fix `ckpt_path="last"` not finding last checkpoints with a templated `CHECKPOINT_NAME_LAST`
#21984 opened 2026-09-30 02:18 by
breken-ai
Fix CSVLogger automatic step restarting at 0 after every flush
#21983 opened 2026-09-30 02:12 by
breken-ai
fix: preserve stateful dataloader state around `estimated_stepping_batches`
#21980 opened 2026-09-29 06:55 by
deependujha
Older