He Ma

@benliebersohn @dlunga The way of finding the model file and model class is a bit of hack right now. See [here](https://github.com/uoguelph-mlrg/Theano-MPI/blob/master/theanompi/worker.py#L179). It requires an absolute path, like something on the...

@Nqabz Thanks for reporting. Let me try debugging it and get back to you.

@Nqabz It seems the Python3 support for Theano-MPI is still experimental but python2 should work since it was developed based on python2. I tried installing [anaconda python3.6](https://www.continuum.io/Downloads) and tested it...

@Nqabz I just tried upgrading my theano to the bleeding-edge version. You just need to change the step 4 in the upper mentioned steps to 4. git clone theano, cd...

@Nqabz Yes. x86_64

Okay. I wil take a look at this when I have time.

@dlunga Ok. Let me run those examples and get back to you.

@dlunga Here are the log files when training Cifar10_model (the toy model): [cifar10-3worker-log.txt](https://github.com/uoguelph-mlrg/Theano-MPI/files/1209676/cifar10-3worker-log.txt) [cifar10-1worker-log.txt](https://github.com/uoguelph-mlrg/Theano-MPI/files/1209677/cifar10-1worker-log.txt) and Wide_ResNet: [wresnet-3worker-log.txt](https://github.com/uoguelph-mlrg/Theano-MPI/files/1209678/wresnet-3worker-log.txt) Single worker Wide_ResNet has not finished yet.

@dlunga The single worker Wide_ResNet log: [wresnet-1worker-log.txt](https://github.com/uoguelph-mlrg/Theano-MPI/files/1209998/wresnet-1worker-log.txt)

@dlunga Note though both dataset classes are named `Cifar10_data`, there are some differences in the class methods between the two for feeding input data. This is due to some Keras...