009e05d14f
[bugfix] Minor improvements to `apply_to_collection` and type signature of `log_dict` ( #7851 )
...
* minor fixeS
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
2021-06-07 09:31:36 +01:00
cfd01d7f8d
move amp checkpoint state management to precision plugin ( #7831 )
2021-06-07 07:45:01 +00:00
dff1047851
Fix an incorrect CHANGELOG link ( #7850 )
2021-06-06 23:57:23 +00:00
7c7182d334
[IPU] Call accelerator hooks regardless if LM hook overridden 1/n ( #7826 )
...
* Modify API to ensure hooks defined in the accelerator are called as expected
* handle step_end in dp
* Add changelog
* Update pytorch_lightning/trainer/trainer.py
Co-authored-by: Carlos Mocholí <carlossmocholi@gmail.com>
* Add todo and explanation
Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>
Co-authored-by: Carlos Mocholí <carlossmocholi@gmail.com>
2021-06-04 16:19:08 +00:00
51d370f4c2
[doc] Move each profiler to its own file + Add missing PyTorchProfiler to the doc ( #7822 )
2021-06-04 21:08:29 +05:30
ca89a7f344
[sharded plugin] Fix check for fp16 precision ( #7825 )
...
Co-authored-by: Carlos Mocholí <carlossmocholi@gmail.com>
Co-authored-by: ananthsub <ananth.subramaniam@gmail.com>
2021-06-04 08:34:39 +02:00
f34584001c
Fix support for torch Module type hints in LightningCLI ( #7807 )
...
* Fixed support for torch Module type hints in LightningCLI
* - Fix issue with serializing values when type hint is Any.
- Run unit test only on newer torchvision versions in which the base class is Module.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Minor change
* Update CHANGELOG.md
Co-authored-by: Carlos Mocholí <carlossmocholi@gmail.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Carlos Mocholí <carlossmocholi@gmail.com>
2021-06-04 07:43:43 +02:00
36770b22fd
validate manual optimization and supported features before running training ( #7788 )
...
Co-authored-by: Carlos Mocholí <carlossmocholi@gmail.com>
2021-06-03 08:42:37 -07:00
03bb389b21
Fix double precision + ddp_spawn ( #6924 )
...
* Initial fix
* Initial fix
* Initial fix
* Updates
* Updates
* Update typing and docs
* Undo accidental refactor
* Remove unused imports
* Add DDP double precision test
* Remove unused variable
* Update CHANGELOG.md
* Fix test
* Update tests
* Formatting
* Revert bad change
* Add back changes
* Correct wrapping order
* Improve unwrapping
* Correct wrapping order
* Fix... finally
* Respond to comments
* Drop ddp test
* Simplify ddp spawn test
* Simplify ddp spawn test
Co-authored-by: Jirka Borovec <Borda@users.noreply.github.com>
2021-06-01 15:21:17 +00:00
195b24ba51
`apply_to_collection` improvements and add `apply_to_collections` ( #7769 )
...
* `apply_to_collection` improvements and add `apply_to_collections`
* Update CHANGELOG
* Minor fix
* Minor fix
* Remove attr
* Swap is first is None
* None test
* OrderedDict support
* flake8
* Fix docstring
2021-06-01 12:09:20 +00:00
1dd61e4e35
Extend support for logging a collection ( #7771 )
2021-06-01 12:51:50 +01:00
0dd6d3a798
Avoid adding `None` loss values in `training_epoch_end` ( #7772 )
2021-05-31 19:28:28 +00:00
7e6010fc93
fix info message when max training time reached ( #7780 )
...
* call time_elapsed
* elapsed formatting
* format
* update test
* changelog
2021-05-31 14:50:16 +02:00
f6b5e3df57
Added save_config_filename init argument to LightningCLI ( #7741 )
2021-05-28 09:30:16 +02:00
9097347ea8
feat(wandb): log models as artifacts ( #6231 )
...
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>
2021-05-27 20:15:02 +02:00
9304c0df8f
Rename and move Result ( #7736 )
...
Co-authored-by: Kaushik B <45285388+kaushikb11@users.noreply.github.com>
2021-05-27 12:27:52 +00:00
906c067b07
Update hooks pseudocode ( #7713 )
2021-05-27 12:27:26 +02:00
04dcb1786d
Add `__len__` method to IndexBatchSamplerWrapper ( #7681 )
...
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
2021-05-26 18:20:13 +02:00
311d9fe67e
Always run validation inside the training loop epoch ( #7357 )
...
Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>
2021-05-26 14:26:48 +02:00
d26953c8bc
Add `ModelPruning(prune_on_train_epoch_end)` to choose when to apply pruning ( #7704 )
...
Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>
2021-05-26 00:57:56 +02:00
7e2f7e956b
fix: improve UserWarning message ( #7685 )
...
* fix: improve UserWarning message
when both overfit and training dtaloader shuffling are enabled
fixes issue: #7656
* chore: update changelog
* Polish userwarning msg in pytorch_lightning/trainer/data_loading.py
Co-authored-by: ananthsub <ananth.subramaniam@gmail.com>
* shuffling typo
* Update CHANGELOG.md
Co-authored-by: ananthsub <ananth.subramaniam@gmail.com>
Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>
Co-authored-by: Carlos Mocholí <carlossmocholi@gmail.com>
2021-05-25 17:35:15 +00:00
e7057d5898
Add `should_rank_save_checkpoint` property to Training Plugins ( #7684 )
...
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
2021-05-25 23:02:05 +05:30
8ba6304c73
Increment the total batch idx before the accumulation early exit ( #7692 )
...
* Increment the total batch idx before the accumulation early exit
* Update CHANGELOG
2021-05-25 10:23:40 +02:00
ad168fc4c6
chlog for 1.3.2 + legacy test ( #7676 )
2021-05-24 17:55:02 +00:00
8b01497e42
Fix global step update when the epoch is skipped ( #7677 )
...
* Fix global step update when the epoch is skipped
* Update CHANGELOG
* Move test
2021-05-24 17:36:56 +01:00
fa41c588f4
Remove ProfilerConnector class ( #7654 )
...
* Remove ProfilerConnector class
* Update trainer.py
* Update CHANGELOG.md
* Update trainer.py
* Update trainer.py
* tests
2021-05-24 08:58:15 -07:00
a54bc5dba3
Fix progress bar print error when called before training ( #7674 )
...
* Check progress bar existence before printing
* Add tests for predict_progres_bar
* Add tests for progress_bar printing without training
* Update changelog
2021-05-24 17:33:28 +02:00
0c958c5a1f
Fix dataloaders are not reset when tuning the model ( #7566 )
...
Co-authored-by: Carlos Mocholi <carlossmocholi@gmail.com>
2021-05-24 10:21:45 +02:00
299f2c481b
FSDP with full state dict ( #7487 )
...
* Fix some test errors
Summary:
Test Plan:
Reviewers:
Subscribers:
Tasks:
Tags:
* checkpoint consolidation
* Update ddp_spawn.py
* Update test_metric_result_integration.py
* Update test_results.py
* Update utils.py
* Update utils.py
* Update test_all_gather_grad.py
* Update test_all_gather_grad.py
* Update test_results.py
* Revert "Update test_results.py"
This reverts commit 9d4a2b891d#1 from shuyingsunshine21/shuyingsunshine21-checkpoint_consolidate"
This reverts commit c5053da7890d23d75bc90d23d75bc970fe5da9c6a9aae99f6eea74906878bf70e431b3f17210183b536c1323b03a9fde915a7a369f47e18222dc98ea6c095b2370250d0aaaa28651d54d79dcdcd29731https://pre-commit.ci
* limit max_epoch to 1 for testing
* test
* fix
* update
* testing remove special for multi gpu
* assert gpu
* add assertion for gpu
* fix
* Re-enable special test, use ModelCheckpoint
* Fix paths
* Fix path passing
* test
* test
* fix test
* fix
* pre-commit format
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
Co-authored-by: Carlos Mocholí <carlossmocholi@gmail.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: SeanNaren <sean@grid.ai>
2021-05-24 08:11:45 +01:00
01109cdf0c
Fix/mismatched toggle optimizer ( #7563 )
...
* fix: avoid potential mismatched toggling of optimzier
Refs #7405
chore: update CHANGELOG
[pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
fix: resolve a confict
chore: update changelog
* feat: add a test that fails in master
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* fix typo in tests/trainer/optimization/test_multiple_optimizers.py
Co-authored-by: ananthsub <ananth.subramaniam@gmail.com>
* Polish tests/trainer/optimization/test_multiple_optimizers.py
Co-authored-by: Carlos Mocholí <carlossmocholi@gmail.com>
* Polish tests/trainer/optimization/test_multiple_optimizers.py
Co-authored-by: Carlos Mocholí <carlossmocholi@gmail.com>
* fix: change placeholder in optimizer_step from positional args to keyword args
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: ananthsub <ananth.subramaniam@gmail.com>
Co-authored-by: Carlos Mocholí <carlossmocholi@gmail.com>
2021-05-23 04:30:28 +02:00
2242423b75
refactor accelerator teardown -> training type plugin teardown ( #7579 )
2021-05-22 13:19:24 -07:00
a8d9b5f783
Remove tbptt `self.log` flags and other dead code [5/n] ( #7644 )
2021-05-22 01:13:00 +00:00
33a1f5271f
[2/N] Define dataclasses for progress tracking ( #7574 )
...
Co-authored-by: ananthsub <ananth.subramaniam@gmail.com>
2021-05-22 03:09:08 +02:00
f6d892ac21
[feat] Support custom filesystems in LightningModule.to_torchscript ( #7617 )
...
* [feat] Support custom filesystems in LightningModule.to_torchscript
* Update CHANGELOG.md
* Update test_torchscript.py
* Update test_torchscript.py
* Update CHANGELOG.md
* Update test_torchscript.py
2021-05-21 11:23:15 +00:00
7eafd8eac6
Add run_name argument to the MLFlowLogger constructor ( #7622 )
...
* Add run_name argument to the MLFlowLogger
* Update CHANGELOG
* Fix unnecessary line
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Fix style by using yapf
* Fix import error when mlflow is not installed
* Update CHANGELOG.md
* Update tests/loggers/test_mlflow.py
Co-authored-by: akiyuki ishikawa <aki.y.ishikwa@gmail.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>
2021-05-21 09:17:32 +01:00
92cf396de2
Override `broadcast_object_list` for `torch<1.8` ( #7592 )
...
Co-authored-by: Carlos Mocholi <carlossmocholi@gmail.com>
2021-05-20 08:29:55 +00:00
ed271905cf
Clear predict_progress_bar in ProgressBar.__getstate__ ( #7608 )
...
Co-authored-by: Yifu Wang <yifuwang@2012@gmail.com>
Co-authored-by: Justus Schock <12886177+justusschock@users.noreply.github.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Carlos Mocholi <carlossmocholi@gmail.com>
2021-05-20 01:38:49 +00:00
8266b141ba
[feat] Support time-based checkpointing during training ( #7515 )
...
Co-authored-by: Jirka Borovec <Borda@users.noreply.github.com>
Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>
2021-05-19 22:14:13 +00:00
9f5d4955b6
[1/N] Define dataclasses for progress tracking ( #6603 )
...
Co-authored-by: Carlos Mocholí <carlossmocholi@gmail.com>
2021-05-19 21:02:20 +00:00
608de6abf4
TensorBoardLogger sub_dir parameter for grouping logs ( #6195 )
...
* fixed a small typo
* cleaning up
* added sub_dir argument to tensorboard and wrote test
* sub dir arg exclusively for tensorboard, linted
* resolving merge conflict
* resolved merge conflict
* resolved merge conflict
* resolved merge conflict
* resolve merge conflict before revert
* resolving merge conflict
* reverted to pre-lint
* added tensorboard sub_dir test
* pep8 formatting
* removed sub_dir arg from test_all function:
* updated feature description
* typo in doc description
* updated CHANGELOG
* Update pytorch_lightning/loggers/tensorboard.py
Co-authored-by: Justus Schock <12886177+justusschock@users.noreply.github.com>
* swapped argument position
* added expandvars tests
* added expandvars
* removed model init
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* fix tests
* fix failed test
* Revert "fix failed test"
This reverts commit 50b34c66da
2021-05-19 19:50:58 +00:00
b4e28e7169
[feat] Add stronger validation for checkpoint_callback argument ( #7539 )
...
* [feat] Add stronger validation for checkpoint_callback configuration
* chlog
* Update callback_connector.py
* Update test_model_checkpoint.py
* Update pytorch_lightning/trainer/connectors/callback_connector.py
Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>
* Update pytorch_lightning/trainer/connectors/callback_connector.py
* Update tests/checkpointing/test_model_checkpoint.py
Co-authored-by: Carlos Mocholí <carlossmocholi@gmail.com>
* Update CHANGELOG.md
Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>
Co-authored-by: Carlos Mocholí <carlossmocholi@gmail.com>
2021-05-19 19:38:08 +00:00
20f63377f8
Fix the condition for calling update_learning_rates ( #7032 )
...
Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
2021-05-17 17:20:42 +02:00
502adbced3
refactor optimizer loop logic for manual and automatic optimization ( #7526 )
...
Co-authored-by: Carlos Mocholí <carlossmocholi@gmail.com>
Co-authored-by: ananthsub <ananth.subramaniam@gmail.com>
Co-authored-by: Justus Schock <12886177+justusschock@users.noreply.github.com>
2021-05-17 14:42:01 +02:00
f4f51e0dcf
Add kubeflow cluster environment ( #7300 )
...
* Add kubeflow cluster environment
* Add KubeflowEnvironment to docs
* Add KubeflowEnvironment to the changelog
* break up a long line
* Add method to detect kubeflow environment
* Select Kubeflow environment when available
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Run pre-commit
* task_idx == 0
Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>
2021-05-17 09:05:24 +01:00
6e6e29af49
remove trainer hidden state | sanity refactor [2 / n] ( #7507 )
2021-05-17 08:57:15 +01:00
d0081778f8
Enable fsspec by default for cli config file ( #7521 )
...
Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>
2021-05-17 08:53:00 +01:00
6ac16ff348
Fix DistribType for `ddp_cpu` (spawn) ( #7492 )
2021-05-14 20:53:26 +01:00
7ca41734da
Add `dataloader_idx` to batch transfer hooks ( #6241 )
...
* replace with kwargs
* chlog
* fix
* add test
* fix
* device
* deepspeed
* pep
* optional
* docs
* bc
* comments
* pep
* mypy
* pep
* Apply suggestions from code review
* kwargs
* docs
* .
* .
* 1.3 -> 1.4
* kwargs -> step_kwargs
2021-05-13 23:03:55 +05:30
a584196abf
Default `seed_everything(workers=True)` in the `LightningCLI` ( #7504 )
2021-05-13 12:18:03 +02:00
dd1a17b071
Refactor result handling in training loop ( #7506 )
...
* refactor results
* rename dic -> dict
* simplify
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* changelog
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* fix None check
* chlog wording
* move process_closure_result to the end
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
2021-05-13 09:30:34 +01:00
298f9e5c2d
Prune deprecated utils modules ( #7503 )
...
* argparse_utils
* model_utils
* warning_utils
* xla_device_utils
* chlog
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
2021-05-13 07:24:42 +00:00
946aee0c7b
prune data parallel ( #7510 )
2021-05-13 06:23:02 +01:00
072ad52b6b
Add `trainer.predict(ckpt_path)` ( #7430 )
...
Co-authored-by: Jirka Borovec <Borda@users.noreply.github.com>
2021-05-13 01:49:58 +02:00
d4ec75164c
Prune deprecated trainer attributes ( #7501 )
...
* use_single_gpu
* use_horovod
* use_ddp2
* use_ddp
* use_dp
* on_gpu
* use_tpu
* on_tpu
* on_cpu
* cleaning
* chlog
* Apply suggestions from code review
* Apply suggestions from code review
2021-05-12 20:10:15 +00:00
96981091c7
Prune deprecated classif. metrics ( #7499 )
...
* stat_scores_multiple_classes
* precision_recall
* precision
* recall
* auc
* auroc
* multiclass_auroc
* iou
* clean-up
* chlog
* flake8
* imports
* prune
2021-05-12 18:03:34 +00:00
140b0c727e
Prune deprecated trainer attributes 2 ( #7502 )
...
* accelerator_backend
* get_model
* clean
* chlog
* flake8
2021-05-12 10:19:30 -07:00
8cdbd03d02
MLFlow now uses env variable as default tracking uri ( #7457 )
...
* Clarify logger flag
Clarify behavior of boolean values on the logger flag for Trainer.
* Update docs/source/common/trainer.rst
* doc
* MLFlow now uses env variable as default tracking uri
Solves https://github.com/PyTorchLightning/pytorch-lightning/issues/6894
* Update pytorch_lightning/loggers/mlflow.py
Co-authored-by: thomas chaton <thomas@grid.ai>
* changelog
Co-authored-by: SpontaneousDuck <kennywitham4@gmail.com>
Co-authored-by: Ethan Harris <ewah1g13@soton.ac.uk>
Co-authored-by: Rohit Gupta <rohitgr1998@gmail.com>
Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>
Co-authored-by: thomas chaton <thomas@grid.ai>
Co-authored-by: jirka <jirka.borovec@seznam.cz>
2021-05-12 11:26:57 +02:00
8538c1f61e
Accelerator model state dict ( #7474 )
...
* Fix some test errors
Summary:
Test Plan:
Reviewers:
Subscribers:
Tasks:
Tags:
* checkpoint consolidation
* Update ddp_spawn.py
* Update test_metric_result_integration.py
* Update test_results.py
* Update utils.py
* Update utils.py
* Update test_all_gather_grad.py
* Update test_all_gather_grad.py
* Update test_results.py
* Revert "Update test_results.py"
This reverts commit 9d4a2b891d#1 from shuyingsunshine21/shuyingsunshine21-checkpoint_consolidate"
This reverts commit c5053da7890d23d75bc90d23d75bc970fe5da9c6a9aae99f6eea74906878bf70e431b3f17210183b536c1323b03a9fde915a7a369f47e18222dc98ea6c095b2370250d0aaaa28651d54d79dcdcd29731https://pre-commit.ci
* Address code review
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: SeanNaren <sean@grid.ai>
2021-05-11 16:39:04 +01:00
7b283e3c46
Bugfix/Multiple dataloaders ( #7433 )
...
* Update supporters.py
* Update apply_func.py
* Update supporters.py
* Update model_train_dataloaders.py
* Update model_train_steps.py
* Update test_dataloaders.py
* Update CHANGELOG.md
* Update model_train_steps.py
* Update test_dataloaders.py
* Update test_dataloaders.py
* Update supporters.py
* Update test_supporters.py
* Apply suggestions from code review
Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Update tests/trainer/test_dataloaders.py
Co-authored-by: Akihiro Nitta <nitta@akihironitta.com>
* Apply suggestions from code review
Co-authored-by: Edgar Riba <edgar.riba@gmail.com>
* Update supporters.py
* Update supporters.py
* Apply suggestions from code review
Co-authored-by: Carlos Mocholí <carlossmocholi@gmail.com>
Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Akihiro Nitta <nitta@akihironitta.com>
Co-authored-by: Edgar Riba <edgar.riba@gmail.com>
Co-authored-by: Carlos Mocholí <carlossmocholi@gmail.com>
2021-05-11 16:33:29 +02:00
d7c44cc649
Docs: sync chlog 1.3.1 ( #7478 )
2021-05-11 12:44:22 +02:00
fdf50a5e4b
Mark certain Trainer APIs as protected ( #7420 )
2021-05-11 11:53:41 +02:00
ad9118f04a
remove trainer hidden state | sanity refactor [1 / n] ( #7437 )
...
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
2021-05-11 11:09:08 +02:00
4a1134db64
Log epoch metrics before firing the `on_evaluation_end` hook ( #7272 )
...
* Log epoch metrics before firing the `on_evaluation_end` hook (addresses #7166 )
* test that epoch metrics are logged before `on_evaluation_end` hook
* update CHANGELOG
* Shorter test
Co-authored-by: Carlos Mocholi <carlossmocholi@gmail.com>
2021-05-11 10:54:31 +02:00
b65ae79478
Automatically check `DataModule.has_{setup,teardown,prepare_data}` [2/2] ( #7238 )
...
* Automatically check `DataModule.has_{setup,teardown,prepare_data}`
* Use variable
* Spacing
* Docs
* Update CHANGELOG
* Remove `_DataModuleWrapper`
* Add test
* Update docs/source/extensions/datamodules.rst
* Bad merge
* add test for invalid name
* Remove ValueError
Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>
2021-05-11 10:53:00 +02:00
987530cd38
Set `num_nodes` and `sync_batchnorm` From Trainer for Manually Passed Training Type Plugin ( #7026 )
...
Co-authored-by: thomas chaton <thomas@grid.ai>
Co-authored-by: Kaushik B <45285388+kaushikb11@users.noreply.github.com>
Co-authored-by: Carlos Mocholi <carlossmocholi@gmail.com>
Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
2021-05-08 11:25:51 +00:00
710b144b9b
Restore `trainer.current_epoch` after tuning ( #7434 )
...
* Add a test
* Save and restore current_epoch
* Update CHANGELOG
* alphabetical order
2021-05-08 07:15:52 +02:00
45143fd825
Improve val step logging ( #7351 )
...
* Fix val step logging
* Add a type
* Fix
* Update CHANGELOG.md
2021-05-07 22:58:03 +00:00
f9e050c5e5
Move DP warning suppression to the DataParallel Plugin ( #7421 )
2021-05-07 23:02:44 +02:00
fecce50355
Deprecate TrainerModelHooksMixin ( #7422 )
...
* Deprecate TrainerModelHooksMixin
* Update CHANGELOG.md
* Update model_hooks.py
* Update model_hooks.py
2021-05-07 13:19:36 -07:00
8208c330eb
Use `torch.nn.utils.clip_grad_norm_` and add `clip_grad_by_value` support for TPU ( #7025 )
...
Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>
Co-authored-by: Kaushik B <45285388+kaushikb11@users.noreply.github.com>
2021-05-07 16:41:39 +00:00
98b94b810c
Fix DeepSpeedPlugin with IterableDataset ( #7362 )
...
* deepspeed add train_micro_batch_size_per_gpu argument
* Update naming and doc
* Modify to use auto naming convention, add test
* Add iterable tests
* Fix tests, attempt by mocking
* Import correct package
* Fix comparison
* Set as special test
* Remove import
* Add Changelog
Co-authored-by: SeanNaren <sean@grid.ai>
2021-05-07 10:46:03 +01:00
28103c67c2
show mush go on ( #7413 )
...
* chlog + version
* readme
* .
2021-05-06 19:06:21 -04:00
fbc8b209f2
update versions ( #7409 )
...
* update versions
* chlog
* win
* str
2021-05-06 20:35:39 +00:00
b181b8c646
release 1.3.0 ( #7404 )
...
* v1.3.0
* ci event
* chlog
* badge
* formatting
2021-05-06 15:05:35 -04:00
d52e0a8f3e
v0.1.3.0rc3 + changelogs ( #7388 )
...
* v0.1.3.0rc3
* spaces
* wip
* wip
* wip
* wip
* prune
* wip
* wip
* Apply suggestions from code review
Co-authored-by: Carlos Mocholí <carlossmocholi@gmail.com>
Co-authored-by: Carlos Mocholí <carlossmocholi@gmail.com>
2021-05-06 07:28:10 -04:00
7b45bcfedb
[2/2] Remove outputs from evaluation epoch end hooks ( #7338 )
...
* Remove outputs from on_train_epoch_end
* iterate
* Update callback_hook.py
* update
* early stop?
* fix
* Update pytorch_lightning/trainer/training_loop.py
Co-authored-by: Ethan Harris <ewah1g13@soton.ac.uk>
* Update trainer.py
* update
* Update training_loop.py
* early stop?
* fix
* Remove outputs from evaluation epoch end hooks
* update
* Update test_remove_1-5.py
* fix lints
* Update base.py
* rm-outputs
* Update evaluation_loop.py
* try-save-more-memory
* Update trainer.py
* Update trainer.py
* cache-at-start
* Update evaluation_loop.py
* Update training_loop.py
* Update training_loop.py
Co-authored-by: Ethan Harris <ewah1g13@soton.ac.uk>
2021-05-05 19:50:58 +00:00
fbcd63aa89
Update changelog for recent releases ( #7387 )
2021-05-05 15:25:56 -04:00
6104a6316a
[1/2] Deprecate `outputs` in `on_train_epoch_end` hooks ( #7339 )
...
* Remove outputs from on_train_epoch_end
* iterate
* Update callback_hook.py
* update
* Update training_loop.py
* Update test_training_loop.py
* early stop?
* fix
* update tests
* Update test_hooks.py
* Update pytorch_lightning/trainer/callback_hook.py
Co-authored-by: Ethan Harris <ewah1g13@soton.ac.uk>
* Update pytorch_lightning/trainer/training_loop.py
Co-authored-by: Ethan Harris <ewah1g13@soton.ac.uk>
* Update trainer.py
* Update pytorch_lightning/trainer/trainer.py
Co-authored-by: Jirka Borovec <Borda@users.noreply.github.com>
Co-authored-by: Ethan Harris <ewah1g13@soton.ac.uk>
Co-authored-by: Jirka Borovec <Borda@users.noreply.github.com>
2021-05-05 17:18:16 +02:00
98670c83a9
Deprecate`truncated_bptt_steps` flag on Trainer in favor of same setting on the LightningModule ( #7323 )
...
* deprecate-tbptt-trainer
* Update CHANGELOG.md
* Update lightning.py
* test
* Update lightning.py
* Update training_loop.py
* Update training_loop.py
* Update lightning.py
* Update training_loop.py
* Update training_loop.py
* update docs
* Update accelerator.py
* Update accelerator.py
* more docs
* tweaks
* chlog
* comments
Co-authored-by: Jirka Borovec <Borda@users.noreply.github.com>
2021-05-05 11:21:00 +01:00
763a9a9495
Fix Namespace loading in PyYAML 5.4.x ( #6673 )
...
* Fix Namespace loading in PyYAML 5.4.x
* Remove OmegaConf reference from PyYAML requirements
* Max allowed version for pyyaml
Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>
2021-05-04 22:56:11 +00:00
374ff750f5
Pass `current_epoch`/`global_step` as monitor candidates [1/2] ( #7344 )
...
* Pass `current_epoch`/`global_step` as monitor candidates
* Formatting
* Fix deprecated test
* Update CHANGELOG
2021-05-04 16:05:40 -04:00
2a740ebe77
Fix support for dataloader with None batches ( #7342 )
...
* Fix Dataloader None batch
* Fix Dataloader None batch
* Update CHANGELOG.md
* Fix breaking test
* Address comments
2021-05-04 12:24:03 +00:00
8c0ea92af2
`TrainerState` refactor [5/5] ( #7173 )
...
* `TrainerState` refactor
* flake8
* Update finished check
* Test cleanup
* Fix tests
* Fixes
* Reorder
* flake8
* Update CHANGELOG
* Better docs
* Better docs
* Remove default
* Update tests
* Bad merge
2021-05-04 12:50:56 +02:00
a6aa1a0f82
make gpus=str in Trainer consistent with command line parsing of string ( #6388 )
...
* string gpu input
* update docs
* deprecation warning
* Revert "update docs"
This reverts commit c5f3893413
2021-05-04 09:56:27 +00:00
2a20102321
fix(wandb): allow custom init args ( #6989 )
...
* feat(wandb): allow custom init args
* style: pep8
* fix: get dict args
* refactor: simplify init args
* test: test init args
* style: pep8
* docs: update CHANGELOG
* test: check default resume value
* fix: default value of anonymous
* fix: respect order of parameters
* feat: use look-up table for anonymous
* yapf formatting
Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>
2021-05-04 09:45:36 +00:00
82c19e1444
Update LR schedulers only when their corresponding Optimizer is being… ( #4868 )
...
* Update LR schedulers only when their corresponding Optimizer is being used.
In the case when optimizer frequencies are specified,
the LR scheduler corresponding to a particular optimizer is updated
only when that optimizer is being used in the training loop or epoch.
* pep8speak fixes
* Fix failing tests
* Add docs
* PR Feedback
* Apply suggestions from code review
Co-authored-by: Rohit Gupta <rohitgr1998@gmail.com>
* formatting fix
* PR Feedback - part 2
* More PR feedback
* Apply suggestions from code review
Co-authored-by: Rohit Gupta <rohitgr1998@gmail.com>
* Add typing imports
* Stronger tests and fixes related to that
* Add more tests plus PR feedback
* Make optimizer_freq_cumsum a cached property
@cached_property is only available after Python 3.8 so had to do it manually.
* Fix tests
* Apply suggestions from code review
Co-authored-by: Carlos Mocholí <carlossmocholi@gmail.com>
* Avoid mutable defaults
* Parametrize lr scheduling tests
* PR feedback
* Apply suggestions from code review
* spell
* Apply suggestions from code review
* flake8
Co-authored-by: Rohit Gupta <rohitgr1998@gmail.com>
Co-authored-by: chaton <thomas@grid.ai>
Co-authored-by: Carlos Mocholí <carlossmocholi@gmail.com>
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
2021-05-04 09:37:40 +00:00
b780af51be
update test for resume_from_checkpoint on missing file ( #7255 )
2021-05-04 09:16:34 +00:00
6da747e775
Deprecate `LightningModule.datamodule` reference in favor of the trainer one ( #6929 ) ( #7168 )
...
Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>
Co-authored-by: Carlos Mocholí <carlossmocholi@gmail.com>
2021-05-04 00:01:41 +00:00
bf1394a472
improve early stopping verbose logging ( #6811 )
2021-05-03 20:20:48 +00:00
14c552bb92
[bugfix] Fix dataloading for iterable datasets and limit_train_batches ( #7306 )
...
* bugfix-dataloading
* rm-logs
* Update CHANGELOG.md
* Update test_dataloaders.py
* Update test_dataloaders.py
* Update training_loop.py
* Update test_dataloaders.py
* Update CHANGELOG.md
* Update CHANGELOG.md
* Update test_dataloaders.py
* Update training_loop.py
* Update training_loop.py
* comments
* address comments
* more tests
* Update progress.py
* Update test_dataloaders.py
* Update test_dataloaders.py
* Update training_loop.py
* Update training_loop.py
* test ckpt fix?
* update again
2021-05-03 19:50:26 +01:00
e0c64f0ef6
Fix Adagrad optimizer not working with DDP/GPU ( #7277 )
...
Co-authored-by: ananthsub <ananth.subramaniam@gmail.com>
Co-authored-by: thomas chaton <thomas@grid.ai>
2021-05-03 03:57:17 +05:30
490cc57809
Device updates for TPU Pod ( #7243 )
2021-04-30 23:14:06 +05:30
16d6c9828d
[bugfix] Apex never instantiated. ( #7274 )
...
* update
* update
* update apex
* update
* update
* update
* remove test.py
* update
* update
* update on comments
* update changelog
* update
* update
* typo
2021-04-30 13:16:28 -04:00
44fd01734c
Move grad_norm to a dedicated utilities file ( #7292 )
...
* rm-grad-norm-mixin
* Update grads.py
* Update CHANGELOG.md
* Apply suggestions from code review
Co-authored-by: Justus Schock <12886177+justusschock@users.noreply.github.com>
Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>
* Update docstrings
* Update __init__.py
Co-authored-by: Carlos Mocholí <carlossmocholi@gmail.com>
Co-authored-by: Justus Schock <12886177+justusschock@users.noreply.github.com>
Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>
2021-04-30 09:19:22 -07:00
e407edba36
[fix] Attach train+val dataloaders to trainer in trainer loop ( #7207 )
...
* Update training_loop.py
* Update test_dataloaders.py
* changelog
* delay reload
* go back
* comments
* Update training_loop.py
* Update test_dataloaders.py
* Update tests/trainer/test_dataloaders.py
Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>
2021-04-30 09:01:31 -07:00
80b9ca0e38
[bugfix] Add reloading support using BaseFinetuning ( #7253 )
...
* update
* wip
* udpate
* update
* update
* update
* resolve bug
* update on comments
* update on comments
* update
* update
* formatting
* add comments
* update on comments
* update
* Update pytorch_lightning/callbacks/base.py
Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>
* update
* update
* Typing and minor changes
* Refactor
* Fix deprecated test
* Broken commit
* Fix broken commit
* flake8
* Update CHANGELOG
* update on comments
Co-authored-by: Jirka Borovec <Borda@users.noreply.github.com>
Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>
Co-authored-by: Carlos Mocholi <carlossmocholi@gmail.com>
2021-04-30 11:14:43 -04:00
5af086ab9f
Attach data refactor and tuner bugs [4/n] ( #7258 )
...
Co-authored-by: Nicki Skafte <skaftenicki@gmail.com>
Co-authored-by: Kaushik B <45285388+kaushikb11@users.noreply.github.com>
2021-04-30 13:54:58 +00:00
b9b3fa371f
fix case where an IterableDataset doesn't produce a batch for an epoch ( #7294 )
...
* wip
* fix
* add test
* refactor + test
* rm
* formatting
* update changelog
* doc
* docstring
* remove unused import
* Update CHANGELOG.md
Co-authored-by: Kaushik B <45285388+kaushikb11@users.noreply.github.com>
Co-authored-by: Kaushik B <45285388+kaushikb11@users.noreply.github.com>
2021-04-30 12:45:55 +00:00
8232de427a
fix save_hyperparameters(container) if container is empty ( #7268 )
...
* fix
* add tests
* changelog
* fix test
2021-04-30 13:38:42 +01:00
338f5a3311
Remove exp_save_path on the LightningModule ( #7266 )
...
* deprecate-exp-save-path
* Update lightning.py
* Update CHANGELOG.md
* remove-not-deprecate
2021-04-29 17:44:04 -04:00
Luis Perez