Alec Graves

I guess I do not understand. This implementation does have a loss added for kl divergence using the mean and log_var / std dev, it is just [added inside of...

Thanks @Gregor0410, I corrected the output of the previous layer to be called logvar.

1. The last commit changed the applied value of beta loss to be summed instead of averaged over the values in the latent space, which I think is what is...

5. Update: the variable named stddev (which was the output of the previous layer) actually represents log variance, which can be negative. I corrected the variable name and undid the...

Yes, the network should adapt to the different range without a problem. Changing the output layer to sigmoid would probably help the network because you are constraining the output to...

This paper has an excellent overview of what the ```beta``` parameter is doing: https://arxiv.org/abs/1804.03599 To summarize, larger ```beta``` will result in a more disentangled latent representation but lower-fidelity reconstructions. Smaller...

It appears to work in later tf versions......

I know it's been a while, but the Theano support branch supports loading the tiny yolo darknet model and retraining.

You will need to manually recreate the model then load the weights. Then you can save the whole model or run predictions, whatever you want. I tried to adress this...

A couple of solutions I have found for allowing both yolo and yolo-tiny to load: First simple, non-robust method using the name of the file: ```python bytes_len = 20 if...