/

/

anima wai

anima wai

AD

About This Model

ANIMA · Base Models

Updated - supports Anima loras! Fine-tune of Anima Base v1 by WAI, which you may know from WAI-Illustrious Since this is a fine-tune of Anima, it should support both tag based prompts, natural language prompts, or a combination of the two. This Workflow supports the following fields: - /scale: Allows upscaling the image during the render process by whichever factor you set here. It will use Ultimate SD Upscale + an anime ESRGAN upscale model - in other words, shiny looks more shiny. - /strength: Sets the denoise value of the upscaling process. You probably won't need to adjust this, but if you find your image is "morphing" during upscale, set this closer to 1.0 (such as 0.7) - /slot1: Setting this to *any* value disables the default positive prompt injection (see notes below). - Also of course supports your typical /size /seed /guidance /steps fields too. Default positive prompt: The author of this fine tune recommends that you start your positive prompt with: > masterpiece, best quality,score_9, score_8, score_7, This is automatically injected into the start of your positive prompt, unless you disable it with the /slot1 setting noted above. The author also recommends starting your negative prompt with: > worst quality, low quality, score_1, score_2, score_3, artist name,blurry, jpeg artifacts, lowres,censor I didn't feel it was super necessarily to add a toggle for negative prompt injection. Model details: https://civitai.red/models/2544636/wai-anima?modelVersionId=2983680