It train on the mix image of latex (latex exp, reflect), i did obverse some unwanted images for me. (Eg, I love the models like Juli and Alex, rather than reflect ones. So, for me, I prefer to update my original model.)

Please use on https://github.com/AUTOMATIC1111/stable-diffusion-webui

Base Model

• Chilloutmix/chilloutmix_NiPrunedFp16Fix.safetensors [59ffe2243a]

• 512 * 768

• Should not have watermark

Tested On

• Chilloutmix/Chilloutmix.safetensors [a757fe8b3d]

Lion optimizer, trained on RTX4090, batch size = 8.

Additional picture source is welcomed, give me some feed back.