Lightning

The core team is excited to announce the PyTorch Lightning 1.6 release ⚡

- [Highlights](highlights)
- [Backward Incompatible Changes](bc-changes)
- [Full Changelog](changelog)
- [Contributors](contributors)

<a name="highlights"></a>
Highlights

PyTorch Lightning 1.6 is the work of 99 contributors who have worked on features, bug-fixes, and documentation for a total of over 750 commits since 1.5. This is our most active release yet. Here are some highlights:

Introducing Intel's Habana Accelerator

Fixed

- Fixed an issue to avoid validation loop run on restart ([11552](https://github.com/PyTorchLightning/pytorch-lightning/pull/11552))
- The Rich progress bar now correctly shows the `on_epoch` logged values on train epoch end ([11689](https://github.com/PyTorchLightning/pytorch-lightning/pull/11689))
- Fixed an issue to make the `step` argument in `WandbLogger.log_image` work ([11716](https://github.com/PyTorchLightning/pytorch-lightning/pull/11716))
- Fixed `restore_optimizers` for mapping states ([11757](https://github.com/PyTorchLightning/pytorch-lightning/pull/11757))
- With `DPStrategy`, the batch is not explicitly moved to the device ([11780](https://github.com/PyTorchLightning/pytorch-lightning/pull/11780))
- Fixed an issue to avoid val bar disappear after `trainer.validate()` ([11700](https://github.com/PyTorchLightning/pytorch-lightning/pull/11700))
- Fixed supporting remote filesystems with `Trainer.weights_save_path` for fault-tolerant training ([11776](https://github.com/PyTorchLightning/pytorch-lightning/pull/11776))
- Fixed check for available modules ([11526](https://github.com/PyTorchLightning/pytorch-lightning/pull/11526))
- Fixed bug where the path for "last" checkpoints was not getting saved correctly which caused newer runs to not remove the previous "last" checkpoint ([11481](https://github.com/PyTorchLightning/pytorch-lightning/pull/11481))
- Fixed bug where the path for best checkpoints was not getting saved correctly when no metric was monitored which caused newer runs to not use the best checkpoint ([11481](https://github.com/PyTorchLightning/pytorch-lightning/pull/11481))

Contributors

ananthsub borda circlecrystal NathanGodey nithinraok rohitgr7


If we forgot someone due to not matching commit email with GitHub account, let us know :]

Fixed

- Pinned sphinx-autodoc-typehints with <v1.15 ([11400](https://github.com/PyTorchLightning/pytorch-lightning/pull/11400))
- Skipped testing with PyTorch 1.7 and Python 3.9 on Ubuntu ([11217](https://github.com/PyTorchLightning/pytorch-lightning/pull/11217))
- Fixed type promotion when tensors of higher category than float are logged ([11401](https://github.com/PyTorchLightning/pytorch-lightning/pull/11401))
- Fixed the format of the configuration saved automatically by the CLI's `SaveConfigCallback` ([11532](https://github.com/PyTorchLightning/pytorch-lightning/pull/11532))

Changed

- Changed `LSFEnvironment` to use `LSB_DJOB_RANKFILE` environment variable instead of `LSB_HOSTS` for determining node rank and main address ([10825](https://github.com/PyTorchLightning/pytorch-lightning/pull/10825))
- Disabled sampler replacement when using `IterableDataset` ([11507](https://github.com/PyTorchLightning/pytorch-lightning/pull/11507))

Contributors

ajtritt akihironitta carmocca rohitgr7

If we forgot someone due to not matching commit email with GitHub account, let us know :]

Fixed

- Fixed `LightningCLI` race condition while saving the config ([11199](https://github.com/PyTorchLightning/pytorch-lightning/pull/11199))
- Fixed the default value used with `log(reduce_fx=min|max)` ([11310](https://github.com/PyTorchLightning/pytorch-lightning/pull/11310))
- Fixed data fetcher selection ([11294](https://github.com/PyTorchLightning/pytorch-lightning/pull/11294))
- Fixed a race condition that could result in incorrect (zero) values being observed in prediction writer callbacks ([11288](https://github.com/PyTorchLightning/pytorch-lightning/pull/11288))
- Fixed dataloaders not getting reloaded the correct amount of times when setting `reload_dataloaders_every_n_epochs` and `check_val_every_n_epoch` ([10948](https://github.com/PyTorchLightning/pytorch-lightning/pull/10948))


Contributors

adamviola akihironitta awaelchli Borda carmocca edpizzi

_If we forgot someone due to not matching commit email with GitHub account, let us know :]_

Fixed

- Fixed `NeptuneLogger` when using DDP ([11030](https://github.com/PyTorchLightning/pytorch-lightning/pull/11030))
- Fixed a bug to disable logging hyperparameters in logger if there are no hparams ([11105](https://github.com/PyTorchLightning/pytorch-lightning/issues/11105))
- Avoid the deprecated `onnx.export(example_outputs=...)` in torch 1.10 ([11116](https://github.com/PyTorchLightning/pytorch-lightning/pull/11116))
- Fixed an issue when torch-scripting a `LightningModule` after training with `Trainer(sync_batchnorm=True)` ([11078](https://github.com/PyTorchLightning/pytorch-lightning/pull/11078))
- Fixed an `AttributeError` occuring when using a `CombinedLoader` (multiple dataloaders) for prediction ([11111](https://github.com/PyTorchLightning/pytorch-lightning/pull/11111))
- Fixed bug where `Trainer(track_grad_norm=..., logger=False)` would fail ([11114](https://github.com/PyTorchLightning/pytorch-lightning/pull/11114))
- Fixed an incorrect warning being produced by the model summary when using `bf16` precision on CPU ([11161](https://github.com/PyTorchLightning/pytorch-lightning/pull/11161))

Changed

- DeepSpeed does not require lightning module zero 3 partitioning ([10655](https://github.com/PyTorchLightning/pytorch-lightning/pull/10655))
- The `ModelCheckpoint` callback now saves and restores attributes `best_k_models`, `kth_best_model_path`, `kth_value`, and `last_model_path` ([10995](https://github.com/PyTorchLightning/pytorch-lightning/pull/10995))


Contributors

awaelchli borchero carmocca guyang3532 kaushikb11 ORippler Raalsky rohitgr7 SeanNaren

_If we forgot someone due to not matching commit email with GitHub account, let us know :]_

Fixed

- Fixed a bug where the DeepSpeedPlugin arguments `cpu_checkpointing` and `contiguous_memory_optimization` were not being forwarded to deepspeed correctly ([10874](https://github.com/PyTorchLightning/pytorch-lightning/issues/10874))
- Fixed an issue with `NeptuneLogger` causing checkpoints to be uploaded with a duplicated file extension ([11015](https://github.com/PyTorchLightning/pytorch-lightning/issues/11015))
- Fixed support for logging within callbacks returned from `LightningModule` ([10991](https://github.com/PyTorchLightning/pytorch-lightning/pull/10991))
- Fixed running sanity check with `RichProgressBar` ([10913](https://github.com/PyTorchLightning/pytorch-lightning/pull/10913))
- Fixed support for `CombinedLoader` while checking for warning raised with eval dataloaders ([10994](https://github.com/PyTorchLightning/pytorch-lightning/pull/10994))
- The TQDM progress bar now correctly shows the `on_epoch` logged values on train epoch end ([11069](https://github.com/PyTorchLightning/pytorch-lightning/pull/11069))
- Fixed bug where the TQDM updated the training progress bar during `trainer.validate` ([11069](https://github.com/PyTorchLightning/pytorch-lightning/pull/11069))


Contributors

carmocca jona-0 kaushikb11 Raalsky rohitgr7

_If we forgot someone due to not matching commit email with GitHub account, let us know :]_