Benjamin Bossan

nolearn uses whatever Lasagne layers you throw at it. I have trained with layers using the `pad` parameter, so I believe that it should work. If you encounter an error...

This looks more like the Lasagne version installed with nolearn not supporting that parameter. Is it possible to instantiate a conv layer with that parameter at all? `layer = layers.ConvLayer(...,...

About your first problem, I don't know why the weights change. nolearn should not interfere with that. Have you tried what happens if you use the same layers but train...

The functions in visualize.py return the plot object. I don't know whether this works, but maybe it's possible to override some attributes and re-draw the images. If this does not...

I see a problem though, since there is not always just one plot. E.g., in [`_plot_head_map`](https://github.com/dnouri/nolearn/blob/master/nolearn/lasagne/visualize.py#L203), there are 3 subplots, which all could have different parameters. Having a general solution...

I was referring to the same problem as mentioned [here](https://groups.google.com/forum/#!topic/lasagne-users/BbG95R6SZ0I). The option you mentioned would not solve this, or would it (can't test it right now)?

I get the same `Cuda not found. Cannot unpickle CudaNdarray` error. Shared variables are saved in Cuda Arrays that you can't load on a machine without Cuda. One suggestion would...

I have a solution that seems to work: ``` python class PortableNeuralNet(NeuralNet): def __setstate__(self, state): # BBB for pickles that don't have the graph with tempfile.TemporaryDirectory() as tmpdirname: filename =...

> But then I found out that this approach has its own problems. Namely, if self.layers is already a list of layer instances Right, I did not think about that...

`sys.setrecursionlimit(10 ** 999)`