| ▲ | trentor 9 hours ago |
| They finally fixed their VAE. It really held back their models over the last 2 years. EDIT: It still produces artifacts it's better but unusable for production work. In midvalues you will see a slight dot pattern. |
|
| ▲ | Ristovski 7 hours ago | parent | next [-] |
| > In midvalues you will see a slight dot pattern. Is this not simply some sort of watermark instead of an artifact? |
| |
| ▲ | trentor 5 hours ago | parent [-] | | No, it's probably their rope implementation. They had a similar problem with the old qwen image but to be sure it needs some digging. |
|
|
| ▲ | mdp2021 8 hours ago | parent | prev [-] |
| > finally fixed their VAE Can you share the sources? |
| |
| ▲ | trentor 8 hours ago | parent [-] | | It's right there in the hugging face link? latents go from 16ch @ 8x compression to 64ch @ 16x, so roughly the same total latent budget but much more channel heavy. It’s also deeper/wider, and the old 2x2 transformer patching is gone. On some images it still produces artifacts but can't say if it's the transformer or the VAE yet. |
|