How can I improve the quality of the generated images? #10
-
|
I don't like the outcome of most of the images. Is there something I must do to further refine them? |
Beta Was this translation helpful? Give feedback.
Answered by
maubreville
Nov 23, 2022
Replies: 1 comment
-
|
I guess there is no really easy answer to your question. In general, Paella has not seen as many images as comparative models (like Stable Diffusion) or has been trained with a similar number of steps. So you essentially can't expect images that are similarly good (as we pointed out in the paper). |
Beta Was this translation helpful? Give feedback.
0 replies
Answer selected by
ClashSAN
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
I guess there is no really easy answer to your question. In general, Paella has not seen as many images as comparative models (like Stable Diffusion) or has been trained with a similar number of steps. So you essentially can't expect images that are similarly good (as we pointed out in the paper).
However, what you could try is to use a model for prompt engineering, such as this one.