SparkTorch: Distributed training of PyTorch networks on Apache Spark with ML Pipeline support

Follow the full discussion on Reddit.
SparkTorch is a project that I have wanted to do for awhile, and after Pytorch released a variety of great updates to the distributed package, I decided to build a package that could easily orchestrate training on Apache Spark. The goal was to be able to easily integrate the training of Pytorch models to the Spark ML Pipeline. This was done by creating a custom estimator that could be saved and loaded for inference (or even additional training). Right now, there are two modes of training: distributed synchronous and Hogwild!. I will be continuing work on the project and would definitely enjoy collaboration.

Comments

There's unfortunately not much to read here yet...

Discover the Best of Machine Learning.

Ever having issues keeping up with everything that's going on in Machine Learning? That's where we help. We're sending out a weekly digest, highlighting the Best of Machine Learning.

Join over 900 Machine Learning Engineers receiving our weekly digest.

Best of Machine LearningBest of Machine Learning

Discover the best guides, books, papers and news in Machine Learning, once per week.

Twitter