Update README.md

README.md

@@ -12,7 +12,7 @@ There are many possible toggles and many variations for this shunt to be used.
 The only one I hooked up is the basic tool meant for simple text encoder guidance, then I shunted it into clip_embeds for a test.
 
 
-Drag and drop into colab and generate some sdxl images with it.
+Drag and drop into colab and generate some sdxl images with it. Two nodes; one above the generator
 
 Fiddle with the taps and mess with the settings to add additional or reduce guidance from the T5-small variations with your clip_l.
 
@@ -181,9 +181,11 @@ config = {
 
 
 gen = torch.Generator(device="cuda").manual_seed(420)
+```
 
+Place this on another node so you don't reload over and over.
 
-
+```
 strength = 0
 
 # the working version that can't be omitted,