Linking xbatcher to tensorflow or pytorch

[dhruvbalwada]

I see in the api reference that there is a reference to pytorch and keras dataloaders. Is there some simple example of how these are actually used?

I am at a stage where I am comfortable using xbatcher to create batches. However, I would like to now be able to do the kind of optimizations that are discussed here, like prefetching and parallelizing. But I am not sure how to move from xbatcher world to tensorflow world.

Any help/examples/suggestions would be great.