Here's how it works in short: Input: Given an image and a written instruction that tells the model what to do Output: The model follows these instructions to edit the image. Checkout the below example:

→ View original post on X — @sumanth_077,