While working on another lora I noticed that Pony doesn't really know the horizontal splits position (it does know vertical splits however).
Versions
1.0: LyCORIS/Dora (requires Automatic1111 1.9+) with extended dataset and improved tagging
0.6: Standard Lora with smaller dataset
Adetailer is recommended for both versions
Tags
The model knows the following 3 main positions (bold are the tags)
horizontal splits
vertical splits
horizontal splits,upside-down
With the following supportive tags you can generate most splits positions:
top-down bottom-up
arched back
upside-down
on stomach
on back
bent over
...
Perspective tags
front view
from behind
from below
from above
from side
...
Negatives
If you notice that the model goes too much 3d, put "3d" in the negative prompt. Some of the images in the dataset leaned a bit too much into 3d for my liking, but I wanted the model to learn the concept better.
"watermark" helps agains the watermark and logos, but unfortunately the model still struggles with it a bit (kinda burnt myself out manually editing the dataset images to get rid of background and watermarks, but as it turned out, this also worsened the model T_T)
I've been told that the model quite frequently seems to create two females instead of one. If that is the case for you, specifying 1girl and/or solo female in positive and 2girls in the negative prompt might help. Most likely this happens because there is one image in the training dataset, which has two females in a vertical pose, because I wanted the model to learn to also support multiple persons.
I will try to fix it in the next version
Resolutions
I mostly use the following resolutions, not sure if others work:
1216x832
1344x768
1024x1024 (sometimes the model tries to fit in the full body making the legs unproportionally short)
768x1344
832x1216
I used a new training method and I'm not entirely content with the results yet, so I will probably retrain the model using my old method, but it will take some time to adjust the dataset etc.
For now enjoy the model, it should work pretty well for most cases
Happy prompting, I'm looking forward to your creations!
Description
-Changed from standard Lora to new Dora type since it seems to perform better
-Changed trigger word from "upside-down splits" to standard "horizontal splits" combined with the "upside-down" tag
-model should have less of an 3D effect
-better prompt adherance
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.