We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
Is there an implementation of the evaluation metrics used in the Dreambooth paper like subject fidelity, prompt fidelity, DINO, CLIP-I and CLIP-T ?