Replies: 1 comment
-
What are the sizes of the images you are trying to generate with the LoRA? I have found that if I generate an image below a certain output resolution, the results are blurry. Usually I need something that is 720x720 or larger. I tend to do 832x1216 or larger and the details improve. |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Hi! Has anyone been able to train a lora model on illustrious to give me advice please?
I have been trying to train a lora model using the illustrious base model without any success.
The final lora only generates noise or very blurry images.
I dont have trouble training an animagine or pony lora, but so far I havent been able to replicate it on Illustrious.
The base for training is the same for every model type, about 20 to 30 images base, txt files with the tagging of the character (And a trigger word), 10 epoch, 10 repeat, 2 ~ 3 batch size. Resulting in a 1500 steps lora aprox.
Thanks! Have a nice day!
Beta Was this translation helpful? Give feedback.
All reactions