search

kubeflow / examples

A repository to host extended examples and tutorials
Apache License 2.0
1.4k stars 753 forks source link

[GSoC] Create a sample for notebook to Kubeflow deployment using TensorFlow 2.0 Keras #774

Closed terrytangyuan closed 4 years ago

terrytangyuan commented 4 years ago

Description: Create a notebook that illustrates the core Kubeflow CUJ:

Reference: The mnist notebooks provide a good example of how to construct an E2E example notebook to be run on Kubeflow and also setup CI/CD for the notebook.

Technical skills required: TensorFlow & Keras, familiarity with Kubeflow and Jupyter

Type: Machine Learning / Data Engineering

Difficulty: Medium

Mentors: Yuan Tang (@terrytangyuan) Ce Gao (@gaocegege)

issue-label-bot[bot] commented 4 years ago

Issue-Label Bot is automatically applying the labels:

Label Probability
kind/feature 0.81

Please mark this comment with :thumbsup: or :thumbsdown: to give our bot feedback! Links: app homepage, dashboard and code for this bot.

kurianbenoy commented 4 years ago

Hi @terrytangyuan, what should we do before GSoC application deadline? Should we send a draft of our GSoC proposal with notebooks showcasing Kubeflow deployment?

What's your specific expectation of a student required to do this task?

terrytangyuan commented 4 years ago

No you don’t have to write notebooks yet. Submitting proposal should be sufficient. Here’s a template you can use to draft your proposal.

siddharthjain1611 commented 4 years ago

What all kubeflow features do you plan for this project?

terrytangyuan commented 4 years ago

@MoAmrYehia Please only ask questions in one place as it's very hard for me to track all the questioins in multiple places. I saw that you've asked on Slack and someone already answered your question.

@siddharthjain1611 Could you be more specific?

siddharthjain1611 commented 4 years ago

@terrytangyuan Sorry to bother you sir, but I got my answers on my proposal. Thanks for replying.

GG-yuki commented 4 years ago

Hi,sir.Sorry to bother you in your busy schedule. I have 2 questions to ask you as following: 1.Should I build a model that can work on something special (high accuracy) ,or I can just use MNIST as an example? 2.I can't create services in Jupyter.There's something confusing in the insrtuction of KFServing,and someone has the same problem in that Issues( maybe because I'm a Rookie hhh).Anyway,I would try to fix this problem or ask in slack. Sincerely,sorry to interrupt you.

11fenil11 commented 4 years ago

@terrytangyuan @gaocegege @ChanYiLin I would love to work on it. As I am working since long with Machine Learning and Jupyter Notebook.

terrytangyuan commented 4 years ago

@GG-yuki To answer your questions:

  1. The model itself is pretty flexible as the main purpose is to show the use of Kubeflow.
  2. Yes, you can post a GitHub issue on KFServing repo or ask on its Slack channel if the instruction is not clear.

@11fenil11 Thanks for your interest. Please submit your proposal as the GSoC deadline for submitting proposals is very close.

11fenil11 commented 4 years ago

I Have shared draft over GSoC website with all my best knowledge about project😊.

Please give comments & feedbacks.

Thanks & Regards. On Mon, 30 Mar, 2020, 6:36 PM Yuan Tang, notifications@github.com wrote:

@GG-yuki https://github.com/GG-yuki To answer your questions:

  1. The model itself is pretty flexible as the main purpose is to show the use of Kubeflow.
  2. Yes, you can post a GitHub issue on KFServing repo or ask on its Slack channel if the instruction is not clear.

@11fenil11 https://github.com/11fenil11 Thanks for your interest. Please submit your proposal as the GSoC deadline https://developers.google.com/open-source/gsoc/timeline for submitting proposals is very close.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/kubeflow/examples/issues/774#issuecomment-605986700, or unsubscribe https://github.com/notifications/unsubscribe-auth/ALHRFJQX2L53C4BYGI6J3ALRKCKLXANCNFSM4LQLY4JQ .

jtfogarty commented 4 years ago

/priority p1

jlewi commented 4 years ago

It might be useful to use and highlight the lineage tracking features with MLMD.

See kubeflow/website#1959

issue-label-bot[bot] commented 4 years ago

Issue-Label Bot is automatically applying the labels:

Label Probability
area/docs 0.90

Please mark this comment with :thumbsup: or :thumbsdown: to give our bot feedback! Links: app homepage, dashboard and code for this bot.

jlewi commented 4 years ago

@terrytangyuan How's this going?

terrytangyuan commented 4 years ago

The student has deployed Kubeflow on GCP and is currently working on the TensorFlow model.

I am linking some issues/PRs he mentioned related to the documentation: https://github.com/kubeflow/website/issues/1985 https://github.com/kubeflow/website/pull/1986 https://github.com/kubeflow/website/issues/1987 https://github.com/kubeflow/website/pull/1988

terrytangyuan commented 4 years ago

@yashjakhotiya Feel free to add anything I am missing here. Also cc @gaocegege @ChanYiLin

stale[bot] commented 4 years ago

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.

terrytangyuan commented 4 years ago

This can be closed now.

/close

k8s-ci-robot commented 4 years ago

@terrytangyuan: Closing this issue.

In response to [this](https://github.com/kubeflow/examples/issues/774#issuecomment-689956590): >This can be closed now. > >/close Instructions for interacting with me using PR comments are available [here](https://git.k8s.io/community/contributors/guide/pull-requests.md). If you have questions or suggestions related to my behavior, please file an issue against the [kubernetes/test-infra](https://github.com/kubernetes/test-infra/issues/new?title=Prow%20issue:) repository.