Dask - ml /Dask -sql with pytorch,keras or tensorflow

1

Is there any documentations or articles with related to how to implement deep learning algorithms using any of these (pytorch,keras and tensorflow) libaries using dask -ml or dask -sql?

I have found documentations with related to implementing deep learning algorithms using skorch and then applying hyperparameter tuning on top of that using dask -ml.But i need to know whether the initial step of applying the model using dask -ml or dask -sql is possible?

Many thanks in Advance

2

Hi @hashanik, welcome to Dask community!

The dask-ml documentation explain what is possible to do with Dask and those libraries:
https://ml.dask.org/pytorch.html

https://ml.dask.org/keras.html

Using dask-ml, you can either learn such a deep learning model on large dataset using Incremental, or perform Hyper parameter search.

Does that answer you question?

Keep in mind that those libraries propose their own set of tools to perform distributed learning.