Lucas Theis

You are right, I should use a tagging library. A switch to django-tagging is planned, but unfortunately I don't have enough time at the moment to actually implement that change.

Hi Christian, I don’t have any immediate ideas but I will try to have a look on the weekend.

Hi Christian, I tried but I wasn't even able to compile cmt with Python 3.8. I guess it is not too surprising since I never properly ported it from Python...

Thanks for all your work. Looks good on first glance. I will need a bit more time to understand why some tests fail before merging.

The units could be clearer in the documentation. The predictions are the expected number of spikes in each bin. The expected number of spikes might be low because of the...

That would be different, yes. The input to the model should to be 100 Hz, because that's what it has been trained on.

Hi Anil, I can't tell you any numbers, but these models (or at least my implementation) are quite memory hungry. I suggest you use the CPU, as the speedup obtained...

Hi Anil, in the paper it's one model trained for each texture. A few years ago we experimented with an MCGSM trained on natural images and only retraining a few...

Hi Anil, to remove 70% of the pixels, I sample from a uniform distribution for each pixel and if a value is less than 0.7, I consider the pixel missing....

Hi Anil, the training script prints the negative log-likelihood, so if it decreases, that's good. I don't have a tutorial for training an MCGSM, only this example: https://github.com/lucastheis/cmt#python-example Lucas