Update README.md

README.md

@@ -1,7 +1,16 @@
 My own version "from scratch" of a self-rescaling CFG / anti-burn. It ain't much but it's honest work.
 
 ## Last update (14.05.24):
-- added node: **preset loader**. Can do what the other can and MUCH MORE! (note: mostly tested with SDXL, the "Red riding latent" preset does not work with SD1.5)). The presets are .json files and can contain a string which will go through eval(). Always check what is inside before running it when it comes from someone else!
+- added node: **preset loader**. Can do what the other can and MUCH MORE!
+    - About some of the presets:
+        - "Enhanced_details_and_tweaked_attention" works better with 24 steps than 12 or on simple compositions.
+        - "Kickstart" acts on input layer 8, it boosts the self attention and then the cross attention when the sigmas are at 7. This tends to create more coherent images.
+        - "The red riding latent" only works with SDXL. It is an almost nonsensical mix of attention tweaks. Best with 12 steps and really nice with creative prompts. Has the tendency to give more red clothings to the characters. Hence the name.
+        - "Excellent_attention" is the default settings for the node described below. Don't delete it or the node won't work.
+        - "Potato Attention Guidance" is really nice for portraits of happy people...
+    - Most of these have been tested on SDXL. I have very little idea of the effect on SD 1.5
+    - The presets are .json files and can contain a string which will go through eval().
+    - Always check what is inside before running it when it comes from someone else!
 - added node: "**Excellent attention**" developped by myself and based on this [astonishingly easy to understand research paper!](https://github.com/Extraltodeus/temp/blob/main/very_science.jpg) But in short:
    - This node allows to disable the input layer 8 on self and cross attention.
    - But also to apply a custom modification on cross attention middle layer 0.