Type | |
Stats | 590 2,800 923 |
Reviews | (142) |
Published | Feb 9, 2025 |
Base Model | |
Training | Steps: 2,000 Epochs: 20 |
Trigger Words | negative space wide shot |
Hash | AutoV2 EA0C8F89D1 |
Negative space
(Read description)
Illustrious version:
About this LoRA:
What is negative space:
According to danbooru:
"Images that are not cropped to focus on their subject, but which include a noticeable amount of seemingly unnecessary empty space in the frame. This might be done to indicate movement, isolation, or simply to leave room for text to be added."
About the model:
This model is not perfect, but it works most of the time on the first attempt. However, sometimes you may need to retry. (Sorry, I'm still a noob at LoRA training, so if you have any advice, I'm interested.) You can increase the weight of this LoRA a bit, but even though I tried to avoid style pollution, there's still a little.
I can't wait to see what you might be able to do with this model.
Note: Don't use portrait because It won't work most of the time, instead use upper body.
How to Use:
Use this LoRA with IllustriousXL or a checkpoint based on this model.
Activation tag: negative space
Weight: 1.0 works fine (again most of the time), you may need to increase this value a little bit.