Update README.md

README.md

@@ -46,15 +46,19 @@ It is an extraordinary compressed model.
 
 This model differs from original "Dark Planet 8B" as follows:
 
-- 12 layers were added to the base models
-- Llama3 instruct was replaced with Llama 3.1 instruct
-- All of the "extended" models (changed from 8b to 10.7B) were "DARE-TIED" together in a framework re-arranging the duplicate layers and replacing these carefully.
+- 12 layers were added to the 8B L3/L3.1 base models, bring them to 10.65 B parameters.
+- Llama 3 instruct was replaced with Llama 3.1 instruct (also extended)
+- All of the "extended" models (changed from 8b to 10.65B) were "DARE-TIED" together in a framework re-arranging the duplicate layers and replacing these carefully.
 
 These changes result in longer output, longer context, and a slight uptick in function of the model.
 
-This is the first version using these extension techniques, with more to follow.
+Content from this model can be especially disturbing, and appear with little warning.
 
-It is for any writing, fiction or roleplay activity.
+IE: "Horror" means real, vivid, and disturbing at times, if you tell the model you want "horror" so to speak.
+
+This is the first version using these extension techniques, with more to follow (already created).
+
+This model is for any writing, fiction or roleplay activity.
 
 It requires Llama3 template and/or "Command-R" template.
 
@@ -73,7 +77,7 @@ Example outputs below.
 - For creative uses, different quants will produce slightly different output.
 - Due to the stability and compressed nature of this model, all quants will operate at above average levels.
 
-This is a LLAMA3.1 model, and requires Llama3 template, but may work with other template(s) and has maximum context of 131k. 
+This is a LLAMA 3.1 model, and requires Llama3 template, but may work with other template(s) and has maximum context of 131k. 
 
 If you use "Command-R" template your output will be very different from using "Llama3" template.