Merge branch 'master' into bugfix/3827_test_ddp_error

Author: tchaton

.github/CODEOWNERS

@@ -12,4 +12,16 @@
 /tests/metrics/*              @teddykoker @ananyahjha93 @justusschock
 /docs/source/metrics.rst      @teddykoker @ananyahjha93 @justusschock
 
+# API
+/pytorch_lightning/callbacks/base.py  @williamfalcon
+/pytorch_lightning/core/datamodule.py @williamfalcon
+/pytorch_lightning/trainer/trainer.py @williamfalcon
+/pytorch_lightning/core/hooks.py     @williamfalcon
+/pytorch_lightning/core/lightning.py  @williamfalcon
 
+
+# accelerators
+/pytorch_lightning/accelerators/*  @williamfalcon
+
+# owners
+/pytorch_lightning/.github/CODEOWNERS @williamfalcon

.github/CONTRIBUTING.md

@@ -112,8 +112,8 @@ In case you adding new dependencies, make sure that they are compatible with the
 
 ### Coding Style
 
-1. Use f-strings for output formation (except logging when we stay with lazy `logging.info("Hello %s!`, name);
-2. Black code formatter is used using `pre-commit` hook.
+1. Use f-strings for output formation (except logging when we stay with lazy `logging.info("Hello %s!", name)`.
+2. Black code formatter is used using a `pre-commit` hook.
 
 ### Documentation
 
@@ -182,10 +182,10 @@ python -m pip install ".[dev, examples]"
 python -m pip install pre-commit
 ```
 
-You can run the full test-case in your terminal via this bash script:
+You can run the full test-case in your terminal via this make script:
 
 ```bash
-bash .run_local_tests.sh
+make test
 ```
 
 Note: if your computer does not have multi-GPU nor TPU these tests are skipped.
@@ -276,6 +276,57 @@ We welcome any useful contribution! For your convenience here's a recommended wo
    git push -f
    ```
 
+4. **How to add new tests**
+
+We are using [pytest](https://docs.pytest.org/en/stable/) in Pytorch Lightning.
+
+Here are tutorials:
+* (recommended) [Visual Testing with pytest](https://www.youtube.com/playlist?list=PLCTHcU1KoD99Rim2tzg-IhYY2iu9FFvNo) from JetBrains on YouTube
+* [Effective Python Testing With Pytest](https://realpython.com/pytest-python-testing/) article on realpython.com
+
+Here is the process to create a new test
+
+* 0. Optional: Follow tutorials !
+* 1. Find a file in tests/ which match what you want to test. If none, create one.
+* 2. Use this template to get started !
+* 3. Use `BoringModel and derivates to test out your code`.
+
+```python
+# TEST SHOULD BE IN YOUR FILE: tests/..../...py
+
+# RUN OUR TEST WITH: pytest tests/..../...py::test_explain_what_is_being_tested --verbose --color=yes --capture=no
+
+# TEST CODE TEMPLATE
+
+# pytest decorator
+# @pytest.mark.skipif(not torch.cuda.is_available(), reason="test requires GPU machine")
+def test_explain_what_is_being_tested(tmpdir):
+    """
+    Test description about text reason to be
+    """
+
+    # os.environ["PL_DEV_DEBUG"] = '1' optional. When activated, you can use internal trainer.dev_debugger
+
+    class ExtendedModel(BoringModel):
+        ...
+
+    model = ExtendedModel()
+
+    # BoringModel is a functional model. You might want to set methods to None to test your behaviour
+    # Example: model.training_step_end = None
+
+    trainer = Trainer(
+        default_root_dir=tmpdir, # will save everything within a tmpdir generated for this test
+        ...
+    )
+    trainer.fit(model)
+    result = trainer.test()
+
+    # assert the behaviour is correct.
+    assert ...
+    assert ...
+```
+
 ### Bonus Workflow Tip
 
 If you don't want to remember all the commands above every time you want to push some code/setup a Lightning Dev environment on a new VM, you can set up bash aliases for some common commands. You can add these to one of your `~/.bashrc`, `~/.zshrc`, or `~/.bash_aliases` files.

.github/workflows/ci_dockers.yml

@@ -66,6 +66,9 @@ jobs:
       fail-fast: false
       matrix:
         include:
+          # todo: see notes in Dockerfile
+          #- python_version: 3.7
+          #  pytorch_version: 1.8
           - python_version: 3.8
             pytorch_version: 1.7
           - python_version: 3.7
@@ -110,9 +113,8 @@ jobs:
             pytorch_version: 1.4
           - python_version: 3.7
             pytorch_version: 1.7
-          # TODO
-          # - python_version: 3.7
-          #   pytorch_version: 1.8
+          - python_version: 3.7
+            pytorch_version: 1.8
     steps:
       - name: Checkout
         uses: actions/checkout@v2

.github/workflows/ci_test-conda.yml

@@ -16,7 +16,7 @@ jobs:
       matrix:
         # os: [ubuntu-20.04]
         python-version: [3.7]
-        pytorch-version: [1.3, 1.4, 1.5, 1.6, 1.7]
+        pytorch-version: [1.3, 1.4, 1.5, 1.6, 1.7, 1.8]
 
     # Timeout: https://stackoverflow.com/a/59076067/4521646
     timeout-minutes: 35

.github/workflows/code-formatting.yml

@@ -7,6 +7,21 @@ on:  # Trigger the workflow on push or pull request, but only for the master bra
     branches: [master, "release/*"]
 
 jobs:
+  imports-check-isort:
+    name: Check valid import formatting with isort
+    runs-on: ubuntu-20.04
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v2
+      - name: Set up Python 3.8
+        uses: actions/setup-python@v2
+        with:
+          python-version: 3.8
+      - name: Install isort
+        run: pip install isort==5.6.4
+      - name: Run isort
+        run: isort --settings-path=./pyproject.toml --check-only --diff .
+
   code-black:
     name: Check code formatting with Black
     runs-on: ubuntu-20.04

.github/workflows/nightly.yml

@@ -70,13 +70,13 @@ jobs:
           tags: pytorchlightning/pytorch_lightning:base-xla-py${{ matrix.python_version }}-torch${{ matrix.xla_version }}
         timeout-minutes: 55
 
-  docker-cuda:
+  docker-cuda-conda:
     runs-on: ubuntu-20.04
     strategy:
       fail-fast: false
       matrix:
         python_version: [3.6, 3.7, 3.8]
-        pytorch_version: [1.3, 1.4, 1.5, 1.6, 1.7]
+        pytorch_version: [1.3, 1.4, 1.5, 1.6, 1.7, 1.8]
         exclude:
           # excludes PT 1.3 as it is missing on pypi
           - python_version: 3.8
@@ -104,6 +104,8 @@ jobs:
         id: extend
 
       - name: Publish CUDA to Docker Hub
+        # ToDo: extend also building for Nightly from pip
+        if: matrix.pytorch_version < 1.8
         # publish master/release
         uses: docker/build-push-action@v2
         with:

.gitignore

@@ -140,3 +140,4 @@ mlruns/
 pytorch\ lightning
 test-reports/
 wandb
+.forked/

.pre-commit-config.yaml

@@ -21,3 +21,12 @@ repos:
     hooks:
       - id: trailing-whitespace
       - id: end-of-file-fixer
+
+  - repo: local
+    hooks:
+      - id: isort
+        name: isort
+        entry: python -m isort
+        args: [--settings-path, ./pyproject.toml]
+        language: system
+        types: [python]

.run_local_tests.sh

@@ -5,7 +5,7 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
 
 
-## [unreleased] - YYYY-MM-DD
+## [unreleased.Features] - YYYY-MM-DD
 
 ### Added
 
@@ -27,21 +27,46 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
 - Added option to log momentum ([#4384](https://github.com/PyTorchLightning/pytorch-lightning/pull/4384))
 
 
-- Added `fsspec` to tuner ([#4458](https://github.com/PyTorchLightning/pytorch-lightning/pull/4458))
+- Added logging using `self.log` in train and evaluation for most callbacks and model hooks (
+    [#4552](https://github.com/PyTorchLightning/pytorch-lightning/pull/4552),
+    [#4495](https://github.com/PyTorchLightning/pytorch-lightning/pull/4495),
+    [#4439](https://github.com/PyTorchLightning/pytorch-lightning/pull/4439))
 
 
-- Added metrics aggregation in Horovod and fixed early stopping ([#3775](https://github.com/PyTorchLightning/pytorch-lightning/pull/3775))
+### Changed
 
+- Tuner algorithms will be skipped if `fast_dev_run=True` ([#3903](https://github.com/PyTorchLightning/pytorch-lightning/pull/3903))
 
-- Added `manual_optimizer_step` which work with `AMP Native` and `accumulated_grad_batches` ([#4485](https://github.com/PyTorchLightning/pytorch-lightning/pull/4485))
 
+- Metric states are no longer as default added to `state_dict` ([#4685](https://github.com/PyTorchLightning/pytorch-lightning/pull/))
+
+
+### Deprecated
+
+
+
+### Removed
+
+
+
+### Fixed
+
+- Allowing decorate model init with saving `hparams` inside ([#4662](https://github.com/PyTorchLightning/pytorch-lightning/pull/4662))
+
+
+- Fixed `setup` callback hook to correctly pass the LightningModule through ([#4608](https://github.com/PyTorchLightning/pytorch-lightning/pull/4608))
+
+
+
+## [unreleased.BugFix] - YYYY-MM-DD
+
+### Added
 
-- Added `persistent(mode)` method to metrics, to enable and disable metric states being added to `state_dict` ([#4482](https://github.com/PyTorchLightning/pytorch-lightning/pull/4482))
 
 
 ### Changed
 
-- Tuner algorithms will be skipped if `fast_dev_run=True` ([#3903](https://github.com/PyTorchLightning/pytorch-lightning/pull/3903))
+
 
 ### Deprecated
 
@@ -53,13 +78,35 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
 
 ### Fixed
 
-- Fixed feature-lack in hpc load ([#4526](https://github.com/PyTorchLightning/pytorch-lightning/pull/4526))
 
 
-- Fixed metrics states being overridden in ddp mode ([#4482](https://github.com/PyTorchLightning/pytorch-lightning/pull/4482))
+## [1.0.6] - 2020-11-11
+
+### Added
+
+- Added metrics aggregation in Horovod and fixed early stopping ([#3775](https://github.com/PyTorchLightning/pytorch-lightning/pull/3775))
+- Added `manual_optimizer_step` which work with `AMP Native` and `accumulated_grad_batches` ([#4485](https://github.com/PyTorchLightning/pytorch-lightning/pull/4485))
+- Added `persistent(mode)` method to metrics, to enable and disable metric states being added to `state_dict` ([#4482](https://github.com/PyTorchLightning/pytorch-lightning/pull/4482))
+- Added congratulations at the end of our notebooks ([#4555](https://github.com/PyTorchLightning/pytorch-lightning/pull/4555))
 
+### Changed
+
+- Changed `fsspec` to tuner ([#4458](https://github.com/PyTorchLightning/pytorch-lightning/pull/4458))
+- Unify SLURM/TorchElastic under backend plugin ([#4578](https://github.com/PyTorchLightning/pytorch-lightning/pull/4578),
+        [#4580](https://github.com/PyTorchLightning/pytorch-lightning/pull/4580),
+        [#4581](https://github.com/PyTorchLightning/pytorch-lightning/pull/4581),
+        [#4582](https://github.com/PyTorchLightning/pytorch-lightning/pull/4582),
+        [#4583](https://github.com/PyTorchLightning/pytorch-lightning/pull/4583))
+
+### Fixed
 
+- Fixed feature-lack in `hpc_load` ([#4526](https://github.com/PyTorchLightning/pytorch-lightning/pull/4526))
+- Fixed metrics states being overridden in DDP mode ([#4482](https://github.com/PyTorchLightning/pytorch-lightning/pull/4482))
 - Fixed `lightning_getattr`, `lightning_hasattr` not finding the correct attributes in datamodule ([#4347](https://github.com/PyTorchLightning/pytorch-lightning/pull/4347))
+- Fixed automatic optimization AMP by `manual_optimization_step` ([#4485](https://github.com/PyTorchLightning/pytorch-lightning/pull/4485))
+- Replace `MisconfigurationException` with warning in `ModelCheckpoint` Callback ([#4560](https://github.com/PyTorchLightning/pytorch-lightning/pull/4560))
+- Fixed logged keys in mlflow logger ([#4412](https://github.com/PyTorchLightning/pytorch-lightning/pull/4412))
+- Fixed `is_picklable` by catching `AttributeError` ([#4508](https://github.com/PyTorchLightning/pytorch-lightning/pull/4508))
 
 
 ## [1.0.5] - 2020-11-03