tencent-ailab / IP-Adapter

The image prompt adapter is designed to enable a pretrained text-to-image diffusion model to generate images with image prompt.
Apache License 2.0
4.46k stars 289 forks source link

difference with the A1111 implementation #359

Open alexblattner opened 1 month ago

alexblattner commented 1 month ago

When you use ip adapter faceid in this repo, it follows the style and prompt of the model much less than the A1111 implementation.

A1111 implementation: image

your implementation: image image

what was done in your implementation that's different?