Katsuya
@kn
If you work with diffusion models and thought the SNR mismatch for the t=T case is odd, someone finally show the impact and propose fixes: https://arxiv.org/pdf/2305.08891.pdf
0 reply
0 recast
0 reaction
