There are many high-quality animation models at present. Considering that the 2D models have not been updated for a long time, the quality of the original training sets is uneven and the image quality is generally blurry. Therefore, the training sets are re-selected and AI is used on some salvageable training images. The image quality is enhanced, the distant view is blurred in the upgraded version of the basic model, and the generalization in color, expression, and environment is significantly improved.Amplification algorithm recommended R-ESRGAN 4x+VAE recommendation: animationvae (character), vae-ft-mse-840000 (scene)CFG ratio: 3-7 (It is recommended to use 3CFG when creating creative compositions, which may produce unexpected effects. The higher the configuration ratio, the more realistic it may be, depending on the prompt, sampler and steps)You can try other settings on your own, and you are welcome to share your creations!

Files with SHA256: 65e813e03663bb8d859daf6620909d300457db0001833f34efff75ead960c45e