3 ms·
I'm curious, in image generation, flow matching is said to be better than diffusion, then why do these language models still start from diffusion, instead of ju
by billconan 1y ago
I'm curious, in image generation, flow matching is said to be better than diffusion, then why do these language models still start from diffusion, instead of jumping to flow matching directly?
- gessha 1y agoThis is just a guess but I think it’s due to diffusion training being more popular so we’ve figured more of the kinks with those models. Flow matching models might follow after you figure out some of their hyperparameters.