Klaus Greff
Klaus Greff
I don't really like `Euclidean` because it doesn't really mean anything in the Neural Networks community. But `CE` is a good suffix to stay consistent. Maybe `GaussianCE` or `RegressionCE` would...
Should we maybe move to a full-blown logging integration? That way we'd get all the configuration options for writing to (rotating) files, filtering the messages, sending via http/smtp and so...
I've added the `SaveLogs` hook that saves all the logs to an HDF5 file. I think this should cover most use cases where some monitors shouldn't be printed but need...
So after giving this a bit more thought, we agree that having a way to direct the printing makes sense for using brainstorm as part of some application. We can...
We can still provide the `__getstate__` and `__setstate__` functions for the usecases that @untom pointed out. But as a standard format for distributing models I agree with @flukeskywalker that hdf5...
Ok, I implemented describing the network and the handlers. From here it should be only a small step to pickleable, and once we've fixed the format for hdf5 that should...
# Format suggestion for HDF5 The simplest format for storing a network as hdf5 would have two groups: - **network**: a JSON serialized version of the network description - **parameters**:...
Ok, I'll just store the JSON encoded as utf-8 as a byte array.
Done for saving and loading the network.
Are we doing anything about this for the release? Currently we can: - store the network (description + parameters as hdf5) - store the trainer logs (using a hook as...