My personal JANIMA workflow for use on ComfyUI, with 2x upscaling included. Simple, (mostly) effective, and easily customizable.
If you actually know what you're doing and notice something wrong or that I can improve, don't be afraid to shoot me a message!
Models and LoRAs used:
https://civitai.red/models/2642932/janima-anima?modelVersionId=2967640 (make sure to also download the dependencies!)
https://civitai.red/models/2638819/748cm-anima-or-artist-styles-or-mexes?modelVersionId=3020719
https://civitai.red/models/2783311/basic-hands-improvement-or-anima?modelVersionId=3134960
https://huggingface.co/Kim2091/2x-AnimeSharpV4/blob/main/2x-AnimeSharpV4_RCAN.safetensors
v1.03:
Added LoRAs ON/OFF Toggle
Added 2 Lora paths, and a toggle between them
Added Anima Turbo LoRA to Iteration Mode
Added AnimeSharp Fix Toggle
Layout redesign
Spaghetti
v1.2:
Added Iteration Mode: simply toggle a button to change between preset step and batch size configurations.
Sharpen Image node already included, located next to the Blur Image node for easy access.
Reorganized the node graph.
v1.1:
Added "Blur Image" node to correct checkerboard artifacting on the final image. Default sigma setting should be good, but you might want to lower it depending on your chosen base resolution. If you want a clarity boost, add a Sharpen Image node after the Upscaler output at 0.4-0.8 alpha instead of lowering the Blur Image sigma below 0.6.
Example Prompts by
https://civitai.red/user/NasuKino (https://civitai.red/images/138581643)
https://civitai.red/user/MagicalAstrogy (https://civitai.red/images/138581643)
Yes, I don't like embedding.