Hi, thanks for sharing your research with us. I am trying to train the diffusion model. I noticed that if labels are used as conditional embedded, then there's an error in conv_blocks.py (emb = emb. reshape(b, c, *((1,)*sp_dim) )) because the embedding has one more dimension (label conditioning). My question is, how do we handle this error?
Hi, thanks for sharing your research with us. I am trying to train the diffusion model. I noticed that if labels are used as conditional embedded, then there's an error in
conv_blocks.py (emb = emb. reshape(b, c, *((1,)*sp_dim) ))because the embedding has one more dimension (label conditioning). My question is, how do we handle this error?