ray/release/nightly_tests/dataset/ray_sgd_training_app.yaml

base_image: "anyscale/ray:nightly-py37-cu102"

python:
  pip_packages: []
  conda_packages: []

post_build_cmds:
    - pip uninstall -y ray
    - pip3 install -U {{ env["RAY_WHEELS"] | default("ray") }}
    - pip3 install llvmlite --ignore-installed
    - pip3 install -U torch==1.6 dask dask_ml mlflow tensorboard torchvision pickle5 s3fs
    - {{ env["RAY_WHEELS_SANITY_CHECK"] | default("echo No Ray wheels sanity check") }}
[CUJ2] add nightly tests for running 500GB ray train (#20195) * add * update cluster env * fix build Co-authored-by: Matthew Deng <matthew.j.deng@gmail.com> 2021-11-21 20:04:45 -08:00			`base_image: "anyscale/ray:nightly-py37-cu102"`

			`python:`
			`pip_packages: []`
			`conda_packages: []`

			`post_build_cmds:`
[Dataset][nighlytest] use latest ray for running test #21148 We are actually using the ray comes with the image, which is on a very old version of Ray. (suprised this actually works) 2021-12-17 23:48:44 -08:00			`- pip uninstall -y ray`
			`- pip3 install -U {{ env["RAY_WHEELS"] \| default("ray") }}`
[CUJ2] add nightly tests for running 500GB ray train (#20195) * add * update cluster env * fix build Co-authored-by: Matthew Deng <matthew.j.deng@gmail.com> 2021-11-21 20:04:45 -08:00			`- pip3 install llvmlite --ignore-installed`
			`- pip3 install -U torch==1.6 dask dask_ml mlflow tensorboard torchvision pickle5 s3fs`
[Dataset][nighlytest] use latest ray for running test #21148 We are actually using the ray comes with the image, which is on a very old version of Ray. (suprised this actually works) 2021-12-17 23:48:44 -08:00			`- {{ env["RAY_WHEELS_SANITY_CHECK"] \| default("echo No Ray wheels sanity check") }}`