* add circleCI
* wip
* CircleCI setup that worked on my private repo. Use a working pytorch-lightning commit
* Fix the orb imports
* Update circleci header comment
* Try to pull the GITHUB_REF from the CI_PULL_REQUEST
* Use null instead of space for 'sed'
* Add TODO for codecov
* Remove echo of GKE_CLUSTER since it will be redacted by CircleCI.
* Try running codecov upload.
* Try using codecov orb
* Use pip install codecov
* Use codecov orb again since it should be approved
* dockers/tpu-tests/Dockerfile
* action
* suggestions
* drop suggestion
* suggestion
Co-authored-by: Jirka <jirka@pytorchlightning.ai>
* Add Github Action to run TPU tests.
* Trigger new Github Actions run.
* Clean up more comments.
* Use different fixed version of ml-testing-accelerators and update config to match.
* use cluster in us-central1-a
* Run 'gcloud logging read' directly without 'echo' to preserve newlines.
* cat coverage.xml on the TPU VM side and upload xml on the Github Action side
* Use new commit on ml-testing-accelerators so command runs fully.
* Preserve newlines in the xml and use if: always() temporarily to upload codecov
* Use pytorch_lightning for coverage instead of pytorch-lightning
* Remove the debug cat of coverage xml
* Apply suggestions from code review
* jsonnet rename
* name
* add codecov flags
* add codecov flags
* codecov
* codecov
* revert codecov
* Clean up after apt-get and remove old TODOs.
* More codefactor cleanups.
* drone
* drone
* disable codecov
* cleaning
* docker py versions
* docker py 3.7
* readme
* bash
* docker
* freeze conda
* py3.6
* Stop using apt-get clean.
* Dont rm pytorch-lightning
* Update docker/tpu/Dockerfile
* Longer timeout in the Github Action to wait for GKE to finish.
* job1
* job2
* job3
Co-authored-by: Jirka Borovec <Borda@users.noreply.github.com>
Co-authored-by: Jirka <jirka@pytorchlightning.ai>
Jirka Borovec