This LoRA focuses on generating sequences related to different kinds of expansion and weight gain.
What's new?
v1.8:
v1.8 has different main trigger words for different sequence lengths:
2 stage sequence: before and after
3 stage sequence: transformation sequence
4 stage sequence: sequential transformation
Using multiple views with 3 or 4 stage sequences might help if the model is being stubborn.
An example prompt for a 2 stage breast expansion might look something like this:
before and after, 1girl, small breasts, huge breasts, breast expansion, ...Or like this if you want three stages:
transformation sequence, 1girl, small breasts, large breasts, huge breasts, breast expansion, ...Recommended resolutions (that I've tested so far):
2 stage sequence: 1024x1024, 1280x1024, 1536x1024, 1368x1024
3 stage sequence: 1368x1024, 1536x1024, 1792x1280
4 stage sequence: 2048x1024
If you end up with boring backgrounds even when specifying a location, try adding this to your negative prompt:
simple background, white background, gray background, gradient backgroundIf your prompt is really complex, the model might struggle to do the correct number of sequence stages.
I don't fully understand my own model yet so I'll update this as I get more experience with it.
Tip
When the start and end sequence look the same, try adding the opposite of what you want earlier in the prompt. Here's an example from the first showcase image:
Positive prompt:
1girl, flat chest, small breasts, long hair, white hair, red eyes, smile, straight hair, holding drink, green bikini,
beach,
breast expansion, gigantic breasts, grabbing own breasts, hand on own breast, skindentation, bursting breastsNegative prompt:
low contrast, ugly, desaturated, low quality, blurry, black and white, grayscale, greyscale, text, watermark, sepiaI start saying flat chest and small breasts before specifying that I want breast expansion and gigantic breasts.
This doesn't always work, but it has been more consistent for me.
This is my first-ever LoRA, so it might be a little rough around the edges as I don't really know what I'm doing...
If you're able to make some good generations with it, feel free to share it with me! I'd love to see what people make.
Description
Tried making the model more consistent by re-tagging the dataset.
FAQ
Looks like we don't have an active mirror for this one right now.
This archive is kept alive by community contributions, so there are still some gaps, especially for newly removed content.
Some files do get recovered over time through contributions. If you're looking for this one, feel free to ask in Discord, or help preserve it if you have a copy.