Questions
#2
by
LOL2024
- opened
Hello, I just have some questions about this model, could you tell me how do you train the model, fine-tune or train from scratch? how many pictures get used for train this model? why architecture is SD1.5 instead of SDXL or any other architectures(such as Flux.1 Schnell or PixArt-Sigma)? And will you add CC0 images from Wikimedia Commons to the training dataset in the future?