Pytorch throws an exception when processing a backward pass from a variable that was obtained from the gaussiansampler function, because exp and add_ are inplace functions. (pytorch version 1.4.0)
Two typos in math, first -- we need to compute KL divergence between posterior and prior distributions, second -- just upper case typo.
Pytorch throws an exception when processing a backward pass from a variable that was obtained from the gaussiansampler function, because exp and add_ are inplace functions. (pytorch version 1.4.0)
Two typos in math, first -- we need to compute KL divergence between posterior and prior distributions, second -- just upper case typo.