This is a Character Lora for Flux based on generated images. An overview of how I train my Lora's can be found here: https://civarchive.com/articles/14453/my-process-for-training-loras
NOTE: This model failed on Epoch 4, I tried out Epoch 3 which is this release and it seems flawless. I guess the bigger the base model, the smaller the lora needs to be.
Description
The training had failed on Z-Image after Epoch 3, so I tried it anyway and it's awesome!. So bottom line, the larger the base model, the smaller the lora needs to be I guess. This release is the Epoch 3 version.