No trigger word needed for this lora. It should impove the quality of images generated for anime and digital illustration overall.
You should pair it with another lora to steer Chroma towards the style you want your image to have, this one should only aims to increase quality.
Trained on 1700 hand picked anime / digital illustration from civitai.
V2 was trained in Onetrainer using Prodigy Plus Schedule Free at batch size 4 for 5000 steps at 512, 768, 1024px
This LoRA is over-fitted since my dataset is too repetitive, it is a good first test, but I will definitely do another version once I have a less repetitive set of images. Batch size also needs to be increased to 8/16 to get better results.
Description
Lora trained at rank 16 on 1000 hand picked high quality images posted on civitai at 1MP, using OneTrainer, with Adamw, -1e04 LR constant, batch size 4, for 11 epochs.
There is a room for improvement by changing the LR to something smaller at the end, adding a reg dataset, ...
The captioning is also mediocre at best (JoyCaption makes a LOT of mistakes).
FAQ
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.