Merge branch 'master' into after-v1.3.6

Author: carmocca

.gitignore

@@ -8,9 +8,7 @@ lightning_logs/
 .vscode/
 
 # Test-tube
-test_tube_logs/
-test_tube_data/
-test_tube_exp/
+test_tube_*/
 
 # Documentations
 docs/source/api

.gitmodules

@@ -1,4 +1,4 @@
-[submodule "notebooks"]
-	path = notebooks
+[submodule "_notebooks"]
+	path = _notebooks
 	url = https://github.com/PyTorchLightning/lightning-tutorials.git
 	branch = publication

.pre-commit-config.yaml

@@ -56,6 +56,7 @@ repos:
     rev: 'v2.3'
     hooks:
       - id: vulture
+        name: Check dead code
 
   - repo: https://github.com/PyCQA/flake8
     rev: 3.9.2

CHANGELOG.md

@@ -9,6 +9,9 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
 
 ### Added
 
+- Add support for named parameter groups in `LearningRateMonitor` ([#7987](https://github.com/PyTorchLightning/pytorch-lightning/pull/7987))
+
+
 - Add `dataclass` support for `pytorch_lightning.utilities.apply_to_collection` ([#7935](https://github.com/PyTorchLightning/pytorch-lightning/pull/7935))
 
 
@@ -77,18 +80,31 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
 - Added IPU Accelerator ([#7867](https://github.com/PyTorchLightning/pytorch-lightning/pull/7867))
 
 
+- Fault-tolerant training
+    * Add `{,load_}state_dict` to `ResultCollection` ([#7948](https://github.com/PyTorchLightning/pytorch-lightning/pull/7948))
+
+
 - Added a warning if `Trainer(log_every_n_steps)` is a value too high for the training dataloader ([#7734](https://github.com/PyTorchLightning/pytorch-lightning/pull/7734))
 
 
 - Added LightningCLI support for argument links applied on instantiation ([#7895](https://github.com/PyTorchLightning/pytorch-lightning/pull/7895))
 
 
+- Added LightningCLI support for configurable callbacks that should always be present ([#7964](https://github.com/PyTorchLightning/pytorch-lightning/pull/7964))
+
+
 - Added DeepSpeed Infinity Support, and updated to DeepSpeed 0.4.0 ([#7234](https://github.com/PyTorchLightning/pytorch-lightning/pull/7234))
 
 
 - Added support for `torch.nn.UninitializedParameter` in `ModelSummary` ([#7642](https://github.com/PyTorchLightning/pytorch-lightning/pull/7642))
 
 
+- Added support `LightningModule.save_hyperparameters` when `LightningModule` is a dataclass ([#7992](https://github.com/PyTorchLightning/pytorch-lightning/pull/7992))
+
+
+- Add support for overriding `optimizer_zero_grad` and `optimizer_step` when using accumulate_grad_batches ([#7980](https://github.com/PyTorchLightning/pytorch-lightning/pull/7980))
+
+
 ### Changed
 
 
@@ -119,8 +135,10 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
     * Moved attributes `hiddens` and `split_idx` to TrainLoop ([#7507](https://github.com/PyTorchLightning/pytorch-lightning/pull/7507))
     * Refactored the logic around manual and automatic optimization inside the optimizer loop ([#7526](https://github.com/PyTorchLightning/pytorch-lightning/pull/7526))
     * Simplified "should run validation" logic ([#7682](https://github.com/PyTorchLightning/pytorch-lightning/pull/7682))
-    * Refactored "should run validation" logic when the trainer is signaled to stop ([#7701](https://github.com/PyTorchLightning/pytorch-lightning/pull/7701))
-
+    * Simplified logic for updating the learning rate for schedulers ([#7682](https://github.com/PyTorchLightning/pytorch-lightning/pull/7682))
+    * Removed the `on_epoch` guard from the "should stop" validation check ([#7701](https://github.com/PyTorchLightning/pytorch-lightning/pull/7701))
+    * Refactored internal loop interface; added new classes `FitLoop`, `TrainingEpochLoop`, `TrainingBatchLoop` ([#7871](https://github.com/PyTorchLightning/pytorch-lightning/pull/7871))
+    * Removed `pytorch_lightning/trainer/training_loop.py` ([#7985](https://github.com/PyTorchLightning/pytorch-lightning/pull/7985))
 
 - Refactored logging
     * Renamed and moved `core/step_result.py` to `trainer/connectors/logger_connector/result.py` ([#7736](https://github.com/PyTorchLightning/pytorch-lightning/pull/7736))
@@ -156,6 +174,8 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
 
 
 - Changed `WandbLogger(log_model={True/'all'})` to log models as artifacts ([#6231](https://github.com/PyTorchLightning/pytorch-lightning/pull/6231))
+
+
 - MLFlowLogger now accepts `run_name` as an constructor argument ([#7622](https://github.com/PyTorchLightning/pytorch-lightning/issues/7622))
 
 
@@ -174,6 +194,12 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
 - Added `on_load_checkpoint` and `on_save_checkpoint` hooks to the `PrecisionPlugin` base class ([#7831](https://github.com/PyTorchLightning/pytorch-lightning/pull/7831))
 
 
+- `LightningCLI` now aborts with a clearer message if config already exists and disables save config during `fast_dev_run`([#7963](https://github.com/PyTorchLightning/pytorch-lightning/pull/7963))
+
+
+- `Trainer(resume_from_checkpoint=...)` now restores the model directly after `LightningModule.setup()`, which is before `LightningModule.configure_sharded_model()` ([#7652](https://github.com/PyTorchLightning/pytorch-lightning/pull/7652))
+
+
 ### Deprecated
 
 
@@ -239,9 +265,18 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
 - Fixed a bug where `precision=64` with `accelerator='ddp_spawn'` would throw a pickle error ([#6924](https://github.com/PyTorchLightning/pytorch-lightning/pull/6924))
 
 
+- Do not override the existing `epoch` value in `logged_metrics` when already logged by the user ([#7982](https://github.com/PyTorchLightning/pytorch-lightning/issues/7982))
+
+
+- Support manual optimization with DeepSpeed ([#7970](https://github.com/PyTorchLightning/pytorch-lightning/pull/7970))
+
+
 - Fixed `dataloader_idx` argument value when predicting with only one `DataLoader` ([#7941](https://github.com/PyTorchLightning/pytorch-lightning/pull/7941))
 
 
+- Pass the `stage` argument of `Callback.{setup,teardown}` as a keyword ([#7973](https://github.com/PyTorchLightning/pytorch-lightning/pull/7973))
+
+
 ## [1.3.6] - 2021-06-15
 
 ### Fixed
@@ -295,6 +330,9 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
 - Fixed training loop total batch counter when accumulate grad batches was enabled ([#7692](https://github.com/PyTorchLightning/pytorch-lightning/pull/7692))
 
 
+- Fixed a bug where skipping an optimizer while using amp causes amp to trigger an assertion error ([#7975](https://github.com/PyTorchLightning/pytorch-lightning/pull/7975))
+
+
 ## [1.3.2] - 2021-05-18
 
 ### Changed

MANIFEST.in

@@ -61,7 +61,7 @@ exclude .pyrightconfig.json
 
 # Exclude submodules
 exclude .gitmodules
-exclude notebooks
+exclude _notebooks
 
 # Exclude Makefile
 exclude Makefile