Type | |
Stats | 1,444 |
Reviews | (209) |
Published | Sep 23, 2023 |
Base Model | |
Hash | AutoV2 7C8520B6CE |
At present, there are many high-quality anime models. Considering that 2D models have not been updated for a long time, the quality of the original training set is uneven and generally blurry. Therefore, a new selection of the training set has been made, and AI image quality has been enhanced for some training images that can be saved. In the upgraded version of the basic model, the foreground is blurry, and the generalization in color, demeanor, and environment is significantly improved.
Recommended amplification algorithm R-ESRGAN 4x+
VAE Recommendation: animevae (character), vae-ft-mse-840000 (scene)
CFG ratio: 3-7 (It is recommended to use 3CFG when creating creative essays, which may produce unexpected effects. The higher the configuration ratio, the more realistic it may be, depending on prompts, samplers, and steps)
You can try other settings yourself, welcome to share your work!