Anime Booster XL LoRA ver 1.0
It enhances the beauty, details and anime style to the images.
Developed on Animagine-XL 3.0 model. TA trainer certified!
Recently, Animmersion XL 1.0 checkpoint was found to work with Anime Booster XL very well. Try it! Check my samples!
Various characters of many animes are trained in Anime Booster XL. If you use the names of some famous animes like kuchiki rukia, inoue orihime, kugisaki nobara, kirishima touka, mikasa, yor briar, yamada asaemon sagiri, etc. (The trigger words of these names are temporarily deleted because they are randomly added to the prompt.), they might show up as familiar faces! Because of the limitation of the number of trigger words, I couldn’t put all the names, you can find out more names. Enjoy~ 🤗🤗🤗 (There are boys too of course~ lol)
Please check the following settings and my postings for example prompts and LoRA combinations:
Use 0.5~1.0 setting depending on your applications. Check out the sample images.
VAE: can be automatic, None or vae-ft-mse-840000-ema-pruned.ckpt. (For Animmersion XL, use sdxl_vae.safetensors.)
Sampling method: Euler, Euler a (You can try others)
sampling steps 25+, cfg 7~10 (With Turbo or LCM checkpoints, it can be much lower)
Adetailer denoising strength: 0.4~0.5
Upscaling: do it as post-processing if needed. (4X-UltraSharp, 4X-AnimeSharp, etc.)
Thanks. Enjoy~
Any advice is appreciated!
** Please follow Tensor Art guideline about the explicit usage of generated images that could result in liability.
