-
Notifications
You must be signed in to change notification settings - Fork 3.4k
Insights: Lightning-AI/pytorch-lightning
Overview
Could not load contribution data
Please try again later
3 Pull requests merged by 3 people
-
update BitsandBytes version
#20313 merged
Oct 2, 2024 -
Add documentation note for TQDMProgressBar
#20198 merged
Sep 30, 2024 -
Make RichProgressBar visible for both light and dark background
#20260 merged
Sep 30, 2024
3 Pull requests opened by 3 people
-
Update version.info to 2.4.1.dev1
#20316 opened
Oct 3, 2024 -
Add learning rate scheduling support for `DeepSpeedStrategy`
#20320 opened
Oct 5, 2024 -
docs: update ref to latest tutorials
#20322 opened
Oct 6, 2024
5 Issues closed by 4 people
-
Unable to serialize WandbLogger
#20315 closed
Oct 5, 2024 -
Smoothing in tqdm progress bar has no effect
#20003 closed
Sep 30, 2024 -
Make `RichProgressBar` visible for both light and dark theme by default
#17118 closed
Sep 30, 2024 -
Validation does not produce any output in PyTorch Lightning using my UNetTestModel
#19751 closed
Sep 30, 2024 -
Split `reload_dataloaders_every_n_epochs` into separate parameters for train, val and test dataloaders
#20309 closed
Sep 29, 2024
5 Issues opened by 5 people
-
best-k-metrics in ModelCheckpoint
#20321 opened
Oct 5, 2024 -
Import error on shutdown/KeyboardInterrupt if ran from Jupyter Lab notebook cell
#20317 opened
Oct 3, 2024 -
Model Checkpointing + FSDP causes Cuda OOM
#20312 opened
Oct 1, 2024 -
Save save_hyperparameters no longer respects linked arguments.
#20311 opened
Sep 30, 2024 -
`hparams` not loaded when loading checkpoint via LightningCLI
#20310 opened
Sep 30, 2024
18 Unresolved conversations
Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.
-
comet_ml logger update
#20275 commented on
Oct 4, 2024 • 3 new comments -
Feature/9947 Add str method to datamodule
#20301 commented on
Oct 3, 2024 • 0 new comments -
Add compile_fn parameter for Trainer
#20269 commented on
Sep 30, 2024 • 0 new comments -
Update LR step scheduler to use total step to work across epochs
#20248 commented on
Sep 30, 2024 • 0 new comments -
Stepwise LR scheduler
#20211 commented on
Sep 30, 2024 • 0 new comments -
fix: correct the positional encoding of Transformer in pytorch examples
#20203 commented on
Sep 30, 2024 • 0 new comments -
feat: allow immutable file upload for wandb logger
#20193 commented on
Sep 30, 2024 • 0 new comments -
Fix: Make `WandbLogger` upload models from all `ModelCheckpoint` callbacks, not just one
#20191 commented on
Sep 30, 2024 • 0 new comments -
make plugin type check more flexible
#20186 commented on
Sep 30, 2024 • 0 new comments -
Add `step` parameter to `TensorBoardLogger.log_hyperparams`
#20176 commented on
Sep 30, 2024 • 0 new comments -
Add email callback on train complete
#20162 commented on
Sep 30, 2024 • 0 new comments -
Loading `train_dataloader` before estimating `max_batches`
#20148 commented on
Sep 30, 2024 • 0 new comments -
Add support to `saving.py` for loading GPU-trained models on CPU-only machines
#19024 commented on
Sep 30, 2024 • 0 new comments -
No action for key "ckpt_path" -> ckpt_path not available for linking
#18885 commented on
Oct 5, 2024 • 0 new comments -
Enable Hyperparameter logging from any hook in the LightningModule
#12624 commented on
Oct 4, 2024 • 0 new comments -
Incosistant memory usage comparing to huggingface trainer when using deepspeed
#20299 commented on
Oct 1, 2024 • 0 new comments -
Support wandb_logger.watch() when using LightningCLI
#19791 commented on
Sep 30, 2024 • 0 new comments -
Registered buffers not moved to correct device when using DeepSpeed Stage 3
#20258 commented on
Sep 29, 2024 • 0 new comments