Using Captions With Dreambooth (JoePenna)
Note : This is now a member-only story. However if you join the Stable Diffusion Training Discord Server, and ask me there (@yushan.777), I will give you a Friends link to view with free access.
Table of Contents
Default Behaviour
1 Level
2 Levels
3 Levels
Multiple Subjects
Captioning When Using Jupyter Notebook (Runpod/Vast.ai)
Information on captioning in JoePenna’s repo can be a bit fragmented. Despite there being a big thread on it on discord, the chatty nature of it can bury important of nuggets of information. In fact the best source of information I found was from the github page of the author of the Captionizer code that is now part of JoePenna’s repo.
This guide assumes you are already familiar and have used JoePenna’s Dreambooth. If not then you can find beginner guides among my other posts (links at the end of this guide)
Default Behaviour
The usual method of training a single subject is to use the token + class pair to represent your subject. So that to invoke your…
