Suggestion to C4W3: TFX on Google Cloud Vertex Pipelines

I’ve successfully completed C4W3 lab “TFX on Google Cloud Vertex Pipelines”. The overall experience is smooth without error.

In the end the model is saved in a GCS bucket. just wonder is it possible for the lab to also include a step to deploy the trained model to an endpoint, so that we students can learn how to deploy in Vertex AI? So far in the lab we’ve learnt how to develop using JupyterLab, to create TFX pipeline, and to select orchestration tool for the runner. It would be great if we can also experience deploying the model, submitting request to endpoint, and seeing the prediction result ourselves!

Thanks again for this interesting lab!

As far as I remember there is part of this course which creates an endpoint for a lab for predictions, at least using docker. Maybe the final week will have that. Also it possible that TD Data and Deployment Specialization includes that.

thanks for your answer. yes I’ve seen endpoint deployment exercise in other labs. but this lab is the only one on Vertex AI platform, just thinking it would be great to have a chance to do endpoint deployment on the AI platform.