Summary:
# Changes
This PR migrates `pytorch_python_doc_build` from circleci to github actions.
Noticeable changes
- Refactor `docker cp` into a single `docker run` with volume mount, because the in circleci volume is not accessible from its remote docker engine
- `pytorch_python_doc_push` job will have a race condition with circleci, which will be migrated in separate PRs
Pull Request resolved: https://github.com/pytorch/pytorch/pull/57371
Reviewed By: samestep
Differential Revision: D28416289
Pulled By: zhouzhuojie
fbshipit-source-id: 04caccccf3d7eb7e2225846a406a53ccda356d44
Summary:
Judging from https://github.com/pytorch/pytorch/issues/57584, it seems like the test-reports artifact was originally intended to be downloaded to `$PWD/test-reports` instead of just directly into `$PWD`. To minimize confusion, this PR changes it to download into `test/test-reports`, which should match where the files came from in the `test` step anyway.
TODOs:
- [x] change the extract path for test-reports
- [x] install Python dependencies
- [x] call `tools/print_test_stats.py`
- [x] use deep clone to allow `git` commands
- [x] correctly set `CIRCLE_*` environment variables
- [x] set Scribe credentials
- [x] set AWS credentials
Pull Request resolved: https://github.com/pytorch/pytorch/pull/57647
Test Plan: CI.
Reviewed By: seemethere
Differential Revision: D28325833
Pulled By: samestep
fbshipit-source-id: cc322bad76747f59b764a1a0a863153bb26095e7
Summary:
This PR adds a shortcut of specifying all models in TorchBench CI
Pull Request resolved: https://github.com/pytorch/pytorch/pull/57311
Test Plan:
CI
RUN_TORCHBENCH: ALL
Reviewed By: bitfort
Differential Revision: D28160198
Pulled By: xuzhao9
fbshipit-source-id: 67c292bc98868979d868d4cf1e599c38e0da94b5
Summary:
This PR adds TorchBench (pytorch/benchmark) CI workflow to pytorch. It tests PRs whose body contains a line staring with "RUN_TORCHBENCH: " followed by a list of torchbench model names. For example, this PR will create a Torchbench job of running pytorch_mobildnet_v3 and yolov3 model.
For security reasons, only the branch on pytorch/pytorch will run. It will not work on forked repositories.
The model names have to match the exact names in pytorch/benchmark/torchbenchmark/models, separated by comma symbol. Only the first line starting with "RUN_TORCHBENCH: " is respected. If nothing is specified after the magic word, no test will run.
Known issues:
1. Build PyTorch from scratch and do not reuse build artifacts from other workflows. This is because GHA migration is still in progress.
2. Currently there is only one worker, so jobs are serialized. We will review the capacity issue after this is deployed.
3. If the user would like to rerun the test, she has to push to the PR. Simply updating the PR body won't work.
4. Only supports environment CUDA 10.2 + python 3.7
RUN_TORCHBENCH: yolov3, pytorch_mobilenet_v3
Pull Request resolved: https://github.com/pytorch/pytorch/pull/56957
Reviewed By: janeyx99
Differential Revision: D28079077
Pulled By: xuzhao9
fbshipit-source-id: e9ea73bdd9f35e650b653009060d477b22174bba
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/57063
Removes the generated tag from the original template so the diff shows
up correctly on internal Phab
Signed-off-by: Eli Uriegas <eliuriegas@fb.com>
Test Plan: Imported from OSS
Reviewed By: H-Huang
Differential Revision: D28040694
Pulled By: seemethere
fbshipit-source-id: c6ec0520fbc4ea169abefc7df2ff925ecc0474cc
Summary:
This way, the user would just have to run the `regenerate_cancel_redundant_workflow.py` script to fix the inconsistency (instead of manual stuff).
Lots of the indentation changes were caused by regenerating the file, which I don't think is terrible, and ruamel.yaml did great at preserving comments and order and such!
Pull Request resolved: https://github.com/pytorch/pytorch/pull/56092
Reviewed By: samestep
Differential Revision: D27780877
Pulled By: janeyx99
fbshipit-source-id: dd2996a88cd70a83d8daac33ba6659f93add8b92
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/56059
The lint doesn't do very much, mostly it enforces that indentation
is consistent. The real point of the lint is to just make sure
that we can still do surgery on codemod with tools like ruamel,
by reusing the configuration in this script.
Signed-off-by: Edward Z. Yang <ezyang@fb.com>
Test Plan: Imported from OSS
Reviewed By: walterddr
Differential Revision: D27774590
Pulled By: ezyang
fbshipit-source-id: c26bc6c95a478bd9b86387b18de7e906e7d13193
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/56056
Since internal systems as well as colab have updated beyond CUDA 10.1
I'd say it's safe to remove CUDA 10.1 builds entirely
As mentioned in https://github.com/pytorch/pytorch/issues/55829#issuecomment-818236019
Signed-off-by: Eli Uriegas <eliuriegas@fb.com>
Test Plan: Imported from OSS
Reviewed By: walterddr
Differential Revision: D27772826
Pulled By: seemethere
fbshipit-source-id: 1599bba26b73b909b2575130219e2708ade5654c
Summary:
This way, the user gets more useful actionable results from the GHA.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/55961
Test Plan: CI
Reviewed By: samestep
Differential Revision: D27749013
Pulled By: janeyx99
fbshipit-source-id: bb0edbcdab29ba8ef99005f6fcf52de6782b468d
Summary:
Currently, we only have three GHA workflows that need to be canceled on reruns. To anticipate for future workflows, this PR enables a check that will make sure any new workflow that should be autocanceled on reruns will be included in the cancel_redundant_workflows.yml GHA workflow.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/55171
Test Plan: Succeeded quick-checks https://github.com/pytorch/pytorch/runs/2249162035?check_suite_focus=true
Reviewed By: samestep
Differential Revision: D27514294
Pulled By: janeyx99
fbshipit-source-id: 27da321f648b97a090052823ec955caffeb6ae97
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/53316
Test Plan:
Nightly Docker build CI
This is a follow-up PR after docker moved default CUDA => 11.1. Only merge this after https://github.com/pytorch/pytorch/issues/53299 is committed.
Reviewed By: albanD
Differential Revision: D26996287
Pulled By: xuzhao9
fbshipit-source-id: 0c2e03da41d036d7aada3e07d479a3dede219f58
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/53133
In light of some issues where users were having trouble installing CUDA
specific versions of pytorch we should no longer have special privileges
for CUDA 10.2.
Recently I added scripts/release/promote/prep_binary_for_pypi.sh (https://github.com/pytorch/pytorch/pull/53056) to make
it so that we could theoretically promote any wheel we publish to
download.pytorch.org to pypi
Signed-off-by: Eli Uriegas <eliuriegas@fb.com>
Test Plan: Imported from OSS
Reviewed By: walterddr
Differential Revision: D26759823
Pulled By: seemethere
fbshipit-source-id: 2d2b29e7fef0f48c23f3c853bdca6144b7c61f22
Summary:
Set CUDA_VERSION to 11.2.0 since Nvidia name their docker image on Ubuntu 18.04 to be nvidia/cuda:11.2.0-cudnn8-devel-ubuntu18.04.
Note that cudatoolkit 11.2.0 is not yet on [conda](https://repo.anaconda.com/pkgs/main/linux-64/), and we need to wait for that before merging this PR.
- https://hub.docker.com/r/nvidia/cuda/
Pull Request resolved: https://github.com/pytorch/pytorch/pull/51990
Reviewed By: samestep
Differential Revision: D26371193
Pulled By: xuzhao9
fbshipit-source-id: 76915490dc30ddb03ceeeadb3c45a6c02b60401e
Summary:
Currently PyTorch repository provides Dockerfile to build Docker with nightly builds, but it doesn't have CI to actually build those Dockers.
This PR adds a GitHub action workflow to create PyTorch nightly build Docker and publish them to GitHub Container Registry.
Also, add "--always" option to the `git describe --tags` command that generates the Docker image tag.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/51755
Test Plan: Manually trigger the workflow build in the GitHub Actions web UI.
Reviewed By: seemethere
Differential Revision: D26320180
Pulled By: xuzhao9
fbshipit-source-id: e00b472df14f5913cab9b06a41e837014e87f1c7
