You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
As we discussed previously: #2021 (comment) we want to add more AI/ML examples to the Kubeflow Training Operator. Right now, most of our examples have very basic and simple CNN training for MNIST. Since Training Operator is capable to train large-scale ML models, we would like to contribute more AI/ML use-cases.
We can make these examples Data Scientists friendly and re-use our Python SDK within Jupyter Notebooks to simplify the user submission.
I like the example structure of HF Transformers, so I propose the following path: examples/<framework>/<ml-use-case>
We can start with these examples (feel free to add more ML use-cases in this issue):
Language Modeling
Image Classification
Text Classification
Audio Classification
Question Answering
Speech Recognition
Text Generation
We should investigate how to configure our CI/CD to make sure that these examples are functional.
As we discussed previously: #2021 (comment) we want to add more AI/ML examples to the Kubeflow Training Operator. Right now, most of our examples have very basic and simple CNN training for MNIST. Since Training Operator is capable to train large-scale ML models, we would like to contribute more AI/ML use-cases.
We can make these examples Data Scientists friendly and re-use our Python SDK within Jupyter Notebooks to simplify the user submission.
I like the example structure of HF Transformers, so I propose the following path: examples/<framework>/<ml-use-case>
We can start with these examples (feel free to add more ML use-cases in this issue):
Language Modeling
Image Classification
Text Classification
Audio Classification
Question Answering
Speech Recognition
Text Generation
We should investigate how to configure our CI/CD to make sure that these examples are functional.
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.
Whenever we publish a new example, please reach out to me or Amber so that we can help turning it into either a short blog post or at least disseminate via social media.
How do you define the actual use case for these topics?
Are these examples supposed to be specific to the training operator or were you thinking of a wider applicability (serving, tuning, metadata, etc.)
As we discussed previously: #2021 (comment) we want to add more AI/ML examples to the Kubeflow Training Operator. Right now, most of our examples have very basic and simple CNN training for MNIST. Since Training Operator is capable to train large-scale ML models, we would like to contribute more AI/ML use-cases.
We can make these examples Data Scientists friendly and re-use our Python SDK within Jupyter Notebooks to simplify the user submission.
I like the example structure of HF Transformers, so I propose the following path:
examples/<framework>/<ml-use-case>
We can start with these examples (feel free to add more ML use-cases in this issue):
We should investigate how to configure our CI/CD to make sure that these examples are functional.
cc @kuizhiqing @johnugeorge @tenzen-y @kubeflow/wg-training-leads
/help
/good-first-issue
/area example
The text was updated successfully, but these errors were encountered: