From 4819ad1c581f63799a79bd7fb99d31a634e7b651 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Carlos=20Mochol=C3=AD?= Date: Thu, 2 Jun 2022 11:19:46 +0200 Subject: [PATCH] Update CHANGELOG after the 1.6.4 release (#13201) --- CHANGELOG.md | 76 +++++++++++++++++++--------------------------------- 1 file changed, 28 insertions(+), 48 deletions(-) diff --git a/CHANGELOG.md b/CHANGELOG.md index 2a2a46302b..c8b2390db1 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -63,9 +63,6 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/). - Added CPU metric tracking to `DeviceStatsMonitor` ([#11795](https://github.com/PyTorchLightning/pytorch-lightning/pull/11795)) -- Added all DDP params to be exposed through hpu parallel strategy ([#13067](https://github.com/PyTorchLightning/pytorch-lightning/pull/13067)) - - - Added `teardown()` method to `Accelerator` ([#11935](https://github.com/PyTorchLightning/pytorch-lightning/pull/11935)) ### Changed @@ -102,12 +99,6 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/). - Changed `pytorch_lightning.core.lightning` to `pytorch_lightning.core.module` ([#12740](https://github.com/PyTorchLightning/pytorch-lightning/pull/12740)) - -- Keep `torch.backends.cudnn.benchmark=False` by default (unlike in v1.6.{0-4}) after speed and memory problems depending on the data used. Please consider tuning `Trainer(benchmark)` manually. ([#13154](https://github.com/PyTorchLightning/pytorch-lightning/pull/13154)) - - -- Prevent modification of `torch.backends.cudnn.benchmark` when `Trainer(benchmark=...)` is not set ([#13154](https://github.com/PyTorchLightning/pytorch-lightning/pull/13154)) - ### Deprecated - Deprecated `pytorch_lightning.loggers.base.LightningLoggerBase` in favor of `pytorch_lightning.loggers.logger.Logger`, and deprecated `pytorch_lightning.loggers.base` in favor of `pytorch_lightning.loggers.logger` ([#120148](https://github.com/PyTorchLightning/pytorch-lightning/pull/12014)) @@ -214,57 +205,46 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/). ### Fixed -- Fixed an issue causing zero-division error for empty dataloaders ([#12885](https://github.com/PyTorchLightning/pytorch-lightning/pull/12885)) - - -- Fixed `Trainer(precision=64)` during evaluation which now uses the wrapped precision module ([#12983](https://github.com/PyTorchLightning/pytorch-lightning/pull/12983)) - - -- Fixed an issue to use wrapped `LightningModule` for evaluation during `trainer.fit` for `BaguaStrategy` ([#12983](https://github.com/PyTorchLightning/pytorch-lightning/pull/12983)) - - -- Fixed mismatching default values for the types of some arguments in the DeepSpeed and Fully-Sharded strategies which made the CLI unable to use them ([#12989](https://github.com/PyTorchLightning/pytorch-lightning/pull/12989)) - - - Fixed an issue with unsupported torch.inference_mode() on hpu backends by making it use no_grad ([#13014](https://github.com/PyTorchLightning/pytorch-lightning/pull/13014)) -- Fixed `DDPStrategy` and `DDPSpawnStrategy` to initialize optimizers only after moving the module to the device ([#11952](https://github.com/PyTorchLightning/pytorch-lightning/pull/11952)) - - -- Fixed epoch logging on train epoch end ([#13025](https://github.com/PyTorchLightning/pytorch-lightning/pull/13025)) - - -- Fixed `materialize_module` setting a module's child recursively ([#12870](https://github.com/PyTorchLightning/pytorch-lightning/pull/12870)) - - -- Fixed the number of references of `LightningModule` so it can be deleted ([#12897](https://github.com/PyTorchLightning/pytorch-lightning/pull/12897)) - - - The model wrapper returned by `LightningLite.setup()` now properly supports pass-through when looking up attributes ([#12597](https://github.com/PyTorchLightning/pytorch-lightning/pull/12597)) -- Avoid redundant callback restore warning while tuning ([#13026](https://github.com/PyTorchLightning/pytorch-lightning/pull/13026)) - - -- Fixed torchelastic detection with non-distributed installations ([#13142](https://github.com/PyTorchLightning/pytorch-lightning/pull/13142)) - - -- Fixed an issue wrt unnecessary usage of habana mixed precision package for fp32 types ([#13028](https://github.com/PyTorchLightning/pytorch-lightning/pull/13028)) - - -- Fixed issue where the CLI could not pass a `Profiler` to the `Trainer` ([#13084](https://github.com/PyTorchLightning/pytorch-lightning/pull/13084)) - - - Fixed issue where the CLI fails with certain torch objects ([#13153](https://github.com/PyTorchLightning/pytorch-lightning/pull/13153)) -- Fixed logging's step values when multiple dataloaders are used during evaluation ([#12184](https://github.com/PyTorchLightning/pytorch-lightning/pull/12184)) - - - +## [1.6.4] - 2022-06-01 + +### Added + +- Added all DDP params to be exposed through hpu parallel strategy ([#13067](https://github.com/PyTorchLightning/pytorch-lightning/pull/13067)) + +### Changed + +- Keep `torch.backends.cudnn.benchmark=False` by default (unlike in v1.6.{0-4}) after speed and memory problems depending on the data used. Please consider tuning `Trainer(benchmark)` manually. ([#13154](https://github.com/PyTorchLightning/pytorch-lightning/pull/13154)) +- Prevent modification of `torch.backends.cudnn.benchmark` when `Trainer(benchmark=...)` is not set ([#13154](https://github.com/PyTorchLightning/pytorch-lightning/pull/13154)) + +### Fixed + +- Fixed an issue causing zero-division error for empty dataloaders ([#12885](https://github.com/PyTorchLightning/pytorch-lightning/pull/12885)) +- Fixed mismatching default values for the types of some arguments in the DeepSpeed and Fully-Sharded strategies which made the CLI unable to use them ([#12989](https://github.com/PyTorchLightning/pytorch-lightning/pull/12989)) +- Avoid redundant callback restore warning while tuning ([#13026](https://github.com/PyTorchLightning/pytorch-lightning/pull/13026)) +- Fixed `Trainer(precision=64)` during evaluation which now uses the wrapped precision module ([#12983](https://github.com/PyTorchLightning/pytorch-lightning/pull/12983)) +- Fixed an issue to use wrapped `LightningModule` for evaluation during `trainer.fit` for `BaguaStrategy` ([#12983](https://github.com/PyTorchLightning/pytorch-lightning/pull/12983)) +- Fixed an issue wrt unnecessary usage of habana mixed precision package for fp32 types ([#13028](https://github.com/PyTorchLightning/pytorch-lightning/pull/13028)) +- Fixed the number of references of `LightningModule` so it can be deleted ([#12897](https://github.com/PyTorchLightning/pytorch-lightning/pull/12897)) +- Fixed `materialize_module` setting a module's child recursively ([#12870](https://github.com/PyTorchLightning/pytorch-lightning/pull/12870)) +- Fixed issue where the CLI could not pass a `Profiler` to the `Trainer` ([#13084](https://github.com/PyTorchLightning/pytorch-lightning/pull/13084)) +- Fixed torchelastic detection with non-distributed installations ([#13142](https://github.com/PyTorchLightning/pytorch-lightning/pull/13142)) +- Fixed logging's step values when multiple dataloaders are used during evaluation ([#12184](https://github.com/PyTorchLightning/pytorch-lightning/pull/12184)) +- Fixed epoch logging on train epoch end ([#13025](https://github.com/PyTorchLightning/pytorch-lightning/pull/13025)) +- Fixed `DDPStrategy` and `DDPSpawnStrategy` to initialize optimizers only after moving the module to the device ([#11952](https://github.com/PyTorchLightning/pytorch-lightning/pull/11952)) + + ## [1.6.3] - 2022-05-03 ### Fixed