Another texual inversion style model from me. This time to get a japanese style image.Similar to my <a rel="ugc" href="https://civarchive.com/models/5131/italian-style">Italian Style TI </a>you can use it to create landscapes as well as portraits or all other kinds of images. No initialization text needed, and the embedding again works on all 1.5 based models.The model is trained on 2000+ images with base 24 base vectors for roughly 2000 steps on my local hardware.

Files with SHA256: 046548aca1b9db87e4ddad9a8359139c5185754b4311844933682feffe0f9d85