All the intermediate checkpoints can be found in <a target="_blank" rel="ugc" href="https://huggingface.co/alea31415/tenten-characters">https://huggingface.co/alea31415/tenten-characters</a>The base model is <a target="_blank" rel="ugc" href="https://huggingface.co/JosephusCheung/ACertainty">ACertainty</a>Trained at clip skip 1.For LoHa you may need to adjust weight according to the base model that you are using, and switch to an earlier checkpoint (<a target="_blank" rel="ugc" href="https://huggingface.co/alea31415/tenten-characters/resolve/main/0318_loha/tenten_loha_step-20000.safetensors">tenten_loha_step-20000.safetensors</a> or <a target="_blank" rel="ugc" href="https://huggingface.co/alea31415/tenten-characters/resolve/main/0318_loha/tenten_loha_step-30000.safetensors">tenten_loha_step-30000.safetensors</a>) may also help.If you find the anime style too strong you can try to put <code>aniscreen</code> in the negative but with a small weight.On the other hand putting <code>aniscreen</code> in the positive recovers the full anime style.Example ImagesLoHa: 2,3,9,10,12,13,14,15,16,18Checkpoint: 1,4,5,6,7,8,11,17,19,20--I don't need support or credit, but I would be glad to know that you are using the models I trained and find it useful.Moreover, I would like to advocate for more franchise models.You can take a look at my workflow <a target="_blank" rel="ugc" href="https://github.com/cyber-meow/anime_screenshot_pipeline">https://github.com/cyber-meow/anime_screenshot_pipeline</a> if you are interested.I just want to spread the fact that there is no reason to encode a single concept in each lora.

Tenten-character-LoHa/FullCkpt 転生王女と天才令嬢の魔法革命

Trained at resolution 512, batch size 8, for 15561 steps

Files with SHA256: b671bbc5684cfb2fc9a1aba5903c1956a18af0ea575de2d7aa00369a3d5cacd2