Open kannon92 opened 11 months ago
I think we should use docker.io/kubeflowkatib/tf-mnist-with-summaries:latest
in the TFJob example since the image is so old.
I can open up a PR to fix that.
@johnugeorge What do you think removing kubeflow/tf-mnist-with-summaries:latest
then using ocker.io/kubeflowkatib/tf-mnist-with-summaries:latest
?
We are building kubeflow/tf-mnist-with-summaries:latest image built each time.
So if update Dockerfile here to use latest tensor flow, won't it be solved automatically?
We are building kubeflow/tf-mnist-with-summaries:latest image built each time.
So if update Dockerfile here to use latest tensor flow, won't it be solved automatically?
I'm not sure if the image works well by just bumping the base image version because we must resolve issues for Python major version bumping (Python 2.7 -> Python 3.x) and Tensorflow major version bumping (TF 1.11 -> TF 2.x).
Newer tensorflow images are not using python 2.x. I just checked tensorflow/tensorflow:2.13.0
and it uses Python 3.8.10
Newer tensorflow images are not using python 2.x. I just checked tensorflow/tensorflow:2.13.0 and it uses Python 3.8.10
I see. I'm ok with bumping TF version if it works without error.
/good-first-issue
@johnugeorge: This request has been marked as suitable for new contributors.
Please ensure the request meets the requirements listed here.
If this request no longer meets these requirements, the label can be removed
by commenting with the /remove-good-first-issue
command.
This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.
/lifecycle frozen
I was using the tensorflow example image to demonstrate JobSet functionality and I saw a lot of warnings related to deprecated features. Not sure if this image will eventually stop working but wanted to bring it to your attention.
Ref: https://github.com/kubernetes-sigs/jobset/pull/253