Jupyter Notebooks

Using Jupyter Notebooks is a great way to serve your ML models, and there are some amazing benefits to using this Kubeflow-hosted component. In this episode of Kubeflow 101, Stephanie Wong will show you how to set up your Jupyter Notebooks on Kubeflow, and speak to the benefits of them - such as better integration with Kubeflow, accessibility of notebooks across teams, integration with Tensorflow, and the ability to leverage GPUs and TPUs.

Product: Kubeflow; fullname: Stephanie Wong;

#Kubeflow101

[starcevda]

Can I connect a 3-gpu rig to the cluster and distribute its resources across multiple notebooks with kubeflow?

[sourabhjain]

But GCP Kubeflow installation does not provide notebook option... It has a separate tab for notebook under AI platform alongside AI Pipelines. If we deploy kubeflow externally, we get the complete kubeflow features as per this video. Could you please provide more inputs on this?

[orjihvy]

but is it to have different notebooks to talk to each other? is that why you need kubernetes for notebooks?

[GaryMillerUK]

one minute in and the question what are jupyter notebooks still hasn't begun to be addressed. I think I'll save myself the next three minutes.