Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Upgrade Python and Pytorch versions for some examples #1906

Merged

Conversation

tenzen-y
Copy link
Member

@tenzen-y tenzen-y commented Jun 26, 2022

What this PR does / why we need it:
I upgraded Python and Pytorch versions for darts-cnn-cifar10 and pytorch-mnist examples.

Which issue(s) this PR fixes (optional, in fixes #<issue number>(, fixes #<issue_number>, ...) format, will close the issue(s) when PR gets merged):
Fixes #1883

Checklist:

  • Docs included if any changes are user facing

@coveralls
Copy link

coveralls commented Jun 26, 2022

Coverage Status

Coverage decreased (-0.1%) to 72.857% when pulling 224a6ef on tenzen-y:upgrade-python-pytorch-version into 9ee8fda on kubeflow:master.

@tenzen-y
Copy link
Member Author

@johnugeorge
This improvement would be better to include in the next Katib release.

Copy link
Member

@anencore94 anencore94 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

/lgtm
It built well in amd64, cpu, linux

@tenzen-y
Copy link
Member Author

I will do force-push to re-run Go Test.

@tenzen-y tenzen-y force-pushed the upgrade-python-pytorch-version branch from dc089b7 to 224a6ef Compare June 27, 2022 12:44
Copy link
Member

@anencore94 anencore94 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

/lgtm

@google-oss-prow google-oss-prow bot added the lgtm label Jun 28, 2022
@johnugeorge
Copy link
Member

/approve

@google-oss-prow
Copy link

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: anencore94, johnugeorge, tenzen-y

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

@google-oss-prow google-oss-prow bot merged commit cfa2d84 into kubeflow:master Jun 28, 2022
@tenzen-y tenzen-y deleted the upgrade-python-pytorch-version branch June 28, 2022 11:50
@djmcgreal-cc
Copy link

Since the base image is no longer built with GPU support, CUDA acceleration is not available. Since we were using this image in our tests, they broke. It's misleading because the mnist.py still tries to use cuda but it will never be available.

@tenzen-y
Copy link
Member Author

Thanks for your comment @djmcgreal-cc.
We have stopped supporting CUDA for the pytorch-mnist image in order to reduce image size due to test-infra issues in #1882.

@johnugeorge
Would you like to provide another pytorch-mnist image with CUDA support?

@johnugeorge
Copy link
Member

@tenzen-y cuda image resources are not available with Github actions?

@tenzen-y
Copy link
Member Author

@tenzen-y cuda image resources are not available with Github actions?

@johnugeorge Since the image with CUDA support size is very heavy as shown below, we will face no space left on device: unknown error.

Not CUDA support image size: 639.63 MB
CUDA support image size: 1.2 GB

@johnugeorge
Copy link
Member

@tenzen-y We can then defer this building step till we find a solution. eg: CI in AWS

Other option, is to manually build and push image as changes are not frequent to this build

@tenzen-y
Copy link
Member Author

@johnugeorge
We can build the image in this workflow, although we can not run an operation check of the image in this workflow.

So, we cannot confirm that the image works, but we can provide the image with support CUDA built in gh-actions.

@johnugeorge
Copy link
Member

@tenzen-y I feel, that should be ok for now

@tenzen-y
Copy link
Member Author

@johnugeorge
I see.
I will create a PR to add the pytorch-mnist-gpu image by the end of this week.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

Upgrade Python and Pytorch version in container image for pytorch-mnist example
5 participants