Why is "-0.5 * torch.sum(1 + sigma - mu.pow(2) - sigma.exp())" in Pytorch equivalent to the KL? proofs variational- ...
確定! 回上一頁