Noah Golmant

Add an analogous example project to pytorch to train resnet on cifar. Refactor accordingly. Add some tests for it. Integrate with the tensorboard stuff when that's done.

This might belong as a track feature. Add integration with tensorboard and tensorboardX for other forms of post-experiment analysis.

Add interaction with ray metric monitoring with track. Add section in README describing how to use ray UI to track experiment progress.

Add feature to load tuples of particular configuration parameters Example config would look like ``` lr, batch_size: grid_search: [(.1, 128), (.2, 256)] ```

Add integration with ray tune to use hyperparameter optimization algorithms like HyperBand.

Long-running code results in `requests` `ChunkedEncodingError` exception (broken connection)

1

I have a script ingesting ~200 GB of landsat imagery with the current multi-threaded implementation (no Dataflow). Eventually, I always get an exception like: ``` requests.exceptions.ChunkedEncodingError: ('Connection broken: IncompleteRead(9186238 bytes...