Art style based on ぁぁぁぁ (@aaaa (Quad-a)) (https://www.pixiv.net/en/users/711048)

v1 Release:

• v1_AnimagineXLV31 is trained on Animagine XL V3.1 (https://civitai.com/models/260267)

• v1_PonyDiffusionV6XL is trained on Pony Diffusion V6 XL (https://civitai.com/models/257749)

• v1_AnythingXL is trained on Anything XL (https://civitai.com/models/9409/or-anything-xl?)

v1 Training Parameters:

• 232 images with Repeats of 7

• Batch size of 5, Epoch of 5

• Learning rate of 0.0003

• Network Rank of 32, Network Alpha of 16

• Clip skip of 2

• No regularisation