Just furry model styles for whatever lolUpdate: the version 1.2 model was trained based on around, 8,000 images and was trained for 3 hours. Most of the art was very much currated to be trained based on a very particular art style. Could I have based the model on 1-3 images. Yeah.. maybe but I think going the extra mile for the last few days. Makes the SDXL lora model that more accurate when doing painting art. Meanwhile, trying to make sure characters look good. When, the LoRa model is used with a checkpoint model. For now unless, I train based on 16,000 images on google colab which might not be possible. I may have to start trying to add a paypal link or something that way if anyone want's to help they are more than welcome. Due, to extreme hardware limitations. don't think this training project can go any further without me getting a much better gpu. Since, the intel pentium gpu really isn't sufficient with ai training. If you want to reach me. You can contact me on discord : zannonkoitsu#0849

Furry styes SDXL LoRA

This model was a bit more tricky. I scavaged different websites for a particular artstyle. An made sure to get around, 2000images. Which I had to manually download -.- so, the model was trained on 1,945 images. And was trained for around, 3 hours. To offer much more detailed painting style that feels more idk expressive without looking to uncanny hopefully. I think in total I did spend around a day or so gathering images over the last few days to help enhance this model. So in reality I think the model was trained on somewhere around, 8,000 images I had to sourt through which was a pain XD since, I had massive hand cramps -.- Because, I doubt this will get used much and for the work I did to make the model as detailed. I hope the 500 points seems reasonable. The v1.3 model will dffiantly be 1,000 points for how much I plan to upgrade the next model.

Files with SHA256: a2095be20d9d49dfe4ada9d457e0f8807208c0f1e2762c6293a9a8c202b07842