Type | |
Stats | 797 945 69 |
Reviews | (93) |
Published | Aug 19, 2023 |
Base Model | |
Training | Steps: 3,500 Epochs: 2 |
Usage Tips | Clip Skip: 1 |
Trigger Words | clayanimation miniature |
Hash | AutoV2 8CBE90F646 |
Here's a LoRa based on clay animation films.
Quite simple to use, it can still sometimes give you strange results... :-)
Warning: don't use negative prompts like bad drawn, bad art, etc. They'll take away everything that makes this LoRa so special.
Works great with Realistic Vision - but also with others ;-)
Warning for SDXL version. This is a Lycoris because I only have a 8go VRAM card and this is the only way to do a training SDXL. And you have to follow this :
Don't use negative prompts like bad drawn, bad art, etc. They'll take away everything that makes this LoRa so special.
Don't use refiner : it's softens the details too much
if the result is "too photorealistic" and doesn't seem to be what's expected, you should add "miniature" to your prompt
Have fun!