* add ppo rl lightning template
* flake
* import gym without try as in qnet example
* fix import format
* remove torch.optim import, not required
* fix import format isort
* add trainer argparse
* change name of trajectory collection method
* add repo in references
* fix typo in comments
* use isinstance to verify actionspace type
* use fstring
* deduplication of logic code
* rename unused forloop variable
* use pl.seed_everything instead
* remove unused numpy import
* format string printed on error
* fix typo in comments
Co-authored-by: chaton <thomas@grid.ai>
* fix num_workers for Windows example
* chlog
* Apply suggestions from code review
Co-authored-by: Carlos Mocholí <carlossmocholi@gmail.com>
* warn
Co-authored-by: Carlos Mocholí <carlossmocholi@gmail.com>
* black formatting and migrated to self.log logging
* Apply suggestions from code review
Co-authored-by: Rohit Gupta <rohitgr1998@gmail.com>
* migrated to accuracy in the metrics package
migrated to accuracy in the metrics package
* removed trailing whitespace
* Apply suggestions from code review
Co-authored-by: Jirka Borovec <Borda@users.noreply.github.com>
Co-authored-by: Rohit Gupta <rohitgr1998@gmail.com>
Co-authored-by: Nicki Skafte <skaftenicki@gmail.com>
(cherry picked from commit 17a0784c5e)
* update DALIClassificationLoader to not use deprecated arguments
* fix line length
* dali version check added and changed args accordingly
* versions
* checking version using disutils.version.LooseVersion now
* .
* ver
* import
Co-authored-by: Jirka Borovec <jirka.borovec@seznam.cz>
Co-authored-by: Jirka Borovec <Borda@users.noreply.github.com>
* add option to step result to do aggregation on a specific device
* in dp: do aggregation on root gpu
* Update CHANGELOG.md
* pep8
* trailing whitespace
* uncomment DP
* more cases
* tmpdir
* test
* note
* move to root
move result
stupid result object
revert to master
undo import
add "to" method to result
generalize to
try a test
try a test
Revert "try a test"
This reverts commit 22e3c1001e6c5774ea18ad925830304c245bf145.
Revert "try a test"
This reverts commit 4d2d8fb2a52d552894809a0cbe51af126d78f070.
new test
max epochs
super epoch end
log in test
hanging test
undo test
initial test that fails on master
step end
pass
step end
step end
epoch end
print
step
check dev
clean up test
sanity check
wtf is go ing on
frustration
debugging test
test
test
test
test
test
test
test
test
unused import
* dist backend -> accelerator
* remove todo
Co-authored-by: Justus Schock <justus.schock@rwth-aachen.de>
Co-authored-by: Justus Schock <12886177+justusschock@users.noreply.github.com>
Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>
* Update to latest logging format and modify the accuracy method.
* Apply suggestions from code review
Co-authored-by: Jirka Borovec <Borda@users.noreply.github.com>
Co-authored-by: chaton <thomas@grid.ai>
Adam's beta 2 parameter was mistakenly referred to as the first order momentum of the gradient, whereas it should be the second order momentum. This has no effect on the correct working of the example.
* add MNIST DALI example, update README.md
* Fix PEP8 warnings
* reformatted using black
* add mnist_dali to test_examples.py
* Add documentation as docstrings
* add nvidia-pyindex and nvidia-dali-cuda100
* replace nvidia-pyindex with --extra-index-url
* mark mnist_dali test as Linux and GPU only
* adjust CUDA docker and examples.txt, fix import error in test_examples.py
* adjust the GPU check
* Exit when DALI is not available
* remove requirements-examples.txt and DALI pip install
* Refactored example, moved to new logging api, added runtime check for test and dali script
* Patch to reflect the mnist example module
* add req.
* Apply suggestions from code review
* Removed requirement as it breaks CPU install, added note in README to install DALI
* add DALI to Drone
* test examples
* Apply suggestions from code review
* imports
* ABC
* cuda
* cuda
* pip DALI
* Move build into init function
Co-authored-by: SeanNaren <sean@grid.ai>
Co-authored-by: Jirka Borovec <jirka@pytorchlightning.ai>
Co-authored-by: Jirka Borovec <Borda@users.noreply.github.com>
Co-authored-by: Sean Naren <sean.narenthiran@gmail.com>
* Removed image generation inside the training step.
It was overwriting the image grid generated in `on_epoch_end`. I also made `adversarial_loss` a static method.
* Incorporated Hyperparameter best practices
Using ArgumentParser and hparams as defined in the Hyperparameters section of
the documentation. This way we can set trainer flags (such as precision,
and gpus) from the command line.
* Incorporated Hyperparameter best practices
Using ArgumentParser and hparams as defined in the Hyperparameters section of
the documentation. This way we can set trainer flags (such as precision,
and gpus) from the command line.
* Split the data part into a LightningDataModule
* Update pl_examples/domain_templates/generative_adversarial_net.py
Co-authored-by: Jeff Yang <ydcjeff@outlook.com>
Jihoon Lee