Update README.md

README.md

@@ -81,3 +81,8 @@ image.save("My_image.png")
 ```
 
 ## Important Notes
+
+1. This model is based on [runwayml/stable-diffusion-v1-5](https://huggingface.co/runwayml/stable-diffusion-v1-5) and this means your older stable diffusion prompts work very well. But consider that it also had different images and text prompts as well. So a little tweaking may seen necessary.
+2. For now, you _must_ put both `blurry` and `low quality` in the _negative prompt_ to get good results.
+3. A big number of inference steps results in a better result, but guidance scale around 10 or 12 is still good. These numbers are there just by experience and you may get different results.
+4. Faces of famous people may be generated with noise or glitch. Consider this when you're generating images of Elon Musk 😂