The new version, trained for SDXL is up.

For this version, I recommend to use weight between 0.8 and 1.5.

It also seems to work with Pony if you use a very high weight, like 3 or 4.

A style LoRA to reproduce classical sketch and drawing study

It's a merge of multiples LoRA trained with the same dataset on SD 1.5 but with different training parameters.

You can use monochrome in your prompt if you dont want colors.

Use a weight of 1.

The training data is also available to download, all the images are in the public domain.