updated the readme file with training settings

README.md

@@ -1,3 +1,14 @@
 ---
 license: creativeml-openrail-m
 ---
+# Arcane Diffusion Dataset
+Dataset containing the 75 images used to train the [Arcane Diffusion](https://huggingface.co/nitrosocke/Arcane-Diffusion) model.
+
+Settings for training:
+```class prompt: illustration style
+instance prompt: illustration arcane style
+learning rate: 5e-6
+lr scheduler: constant
+num class images: 1000
+max train steps: 5000
+```