This lora is an experiment, by mixing several subjects in the training dataset, I have created a new person! Just prompt different hair color and other details and Viola!
Trained on 37 carefully chosen pics, for 5 epochs and approximately 1850 steps at a batch of 2.
Dim was set at 32/8
Experiment with steps and sampling method. If using hires.fix, use low values.
Description
Try some different prompts and lets see those results!
Details
Downloads
7
Platform
SeaArt
Platform Status
Available
Created
5/17/2024
Updated
5/17/2024
Deleted
-
Trigger Words:
M3tz







