search

YuxinWenRick / hard-prompts-made-easy

MIT License
601 stars 54 forks source link

[Discussion] Difference between different models? And other values? #23

Closed StableInfo closed 1 year ago

StableInfo commented 1 year ago

Hello, How to know what is the best model to choose from when trying to match the wanted prompt with the models we usually use (huggingface, civitai..) ?

YuxinWenRick commented 1 year ago

Hi, thanks for your interest!

Did you mean which CLIP model you should use for the prompt optimization?

There might not be a straightforward way to do so, but you can first check out the huggingface page of the diffusion model you are using to see what CLIP model they are using. For example, for runwayml/stable-diffusion-v1-5, if you search for "text encoder," you will see that they are using CLIP ViT-L/14.

After you check the version of the CLIP model they used, you can modify the hyperparameter accordingly, as discussed here and here.