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
I have been unable to train lora models since a recent windows re-install. Stable diffusion in automatic1111 is working fine and all models created prior work perfectly. However, when I attempt to train a new model it appears to work and goes through the entire process but the resulting model is not actually trained. When I attempt to use it, the results are not at all close to the training data. A model trained on a person will return a picture of a house or a field. If it does return a person it looks nothing like the person it is supposed to. The one thing I did notice is that the resulting file is VERY small. 9MB instead of 145MB. It looks like it is creating the model but not actually training it. I also just discovered that if I go back to dreambooth in Automatic1111 I can't train there either. It won't even let me choose xformers which was the only way I was previously able to get lora training to work there. Any ideas?
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
-
I have been unable to train lora models since a recent windows re-install. Stable diffusion in automatic1111 is working fine and all models created prior work perfectly. However, when I attempt to train a new model it appears to work and goes through the entire process but the resulting model is not actually trained. When I attempt to use it, the results are not at all close to the training data. A model trained on a person will return a picture of a house or a field. If it does return a person it looks nothing like the person it is supposed to. The one thing I did notice is that the resulting file is VERY small. 9MB instead of 145MB. It looks like it is creating the model but not actually training it. I also just discovered that if I go back to dreambooth in Automatic1111 I can't train there either. It won't even let me choose xformers which was the only way I was previously able to get lora training to work there. Any ideas?
Beta Was this translation helpful? Give feedback.
All reactions