Change default model
#15
by
YuLexuan30
- opened
The default model for training is Stable Diffusion 2.1, which is known to be bad and nobody uses. Change it to SDXL 1.0 or allow users to specify another model of their choice.