Lora trained in recreate Amy Matthews' art style (or CartoonPink).

Thank you so much to MarasC! Who provided the dataset and help me a lot to prune and clean it. This model could not have been created without their help (>°-°)>,

This Lora was a nightmare, it took around of 16 hours for clean and prune the whole dataset, plus 22 hours of training (4 retries of 5.5 hours each one to make it). There is still space for improvement but I don't want retrain it by 5° time in a row :panic:.

Version 1:

You can use this Lora from the weight .4 to 1, (I added several examples about how changes the results). Has some features from the artist like the thick eyebrows or prominent lips and nipples, you can reduce them with the next negatives. The traditional media is just a helper for make the style more alike but isn't required, I prefer don't use it.

Aux tags: traditional media

Negative Aux tags: thick eyebrows, lips, erect nipples

Weight: .4 - 1