You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When I was training the lora in Kohya_ss, I noticed 2 parameters Min Timestep and Max Timestep. Judging by the description, they somehow affect the work of lora in img2img mode. I'm interested in this point because I want to create a lora style focused more on img2img to convert certain images to the style I need. All this is also problematic because these points are hardly mentioned on the Internet and in the analysis of training parameters.
How do these parameters work? What is the best value I can set for them and do I need to set any other checkboxes or values for them?
The text was updated successfully, but these errors were encountered:
When I was training the lora in Kohya_ss, I noticed 2 parameters Min Timestep and Max Timestep. Judging by the description, they somehow affect the work of lora in img2img mode. I'm interested in this point because I want to create a lora style focused more on img2img to convert certain images to the style I need. All this is also problematic because these points are hardly mentioned on the Internet and in the analysis of training parameters.
How do these parameters work? What is the best value I can set for them and do I need to set any other checkboxes or values for them?
The text was updated successfully, but these errors were encountered: