Update README.md

README.md

@@ -40,7 +40,7 @@ pipeline_tag: text-generation
 
 <img src="pcuc.jpg" style="float:right; width:300px; height:300px; padding:10px;">
 
-This is a specialized prose/creative model based in Llama2, max context of 4096 (or 16k+ with rope).
+This is a specialized prose / creative model based in Llama2, max context of 4096 (or 16k+ with rope).
 
 This model has been designed to be relatively bullet proof and operates with most parameters, including temp settings from 0 to 5.
 
@@ -150,13 +150,18 @@ The process to modify the model occurs at the root level - source files level. T
 This is a slightly experimental model, and as a result it may "glitch" from time to time - the most common is
 it may mispell a word (or two) during generation.
 
-So if you are going to use this model for any type of generation, spell check / check in WORD etc etc just in case.
+Adjusting rep pen and/or temp many times can fix this issue.
+
+Likewise more detailed prompts seem to lessen or prevent this issue too.
+
+If you are going to use this model for any type of generation, spell check / check in WORD etc etc just in case.
 
 Also note, this is a Llama2 model and the brainstorm process pushes right to the wall in terms of performance.
 
 The more detailed your prompt, the better the model will perform.
 
-Llama 3, 3.1, Mistral, Mistral Nemo, Solar models performance is stronger than LLama2's.
+Llama 3, 3.1, Mistral, Mistral Nemo, Solar models performance is stronger than LLama2's... yet the examples 
+show just what "old" Llama2s can do.
 
 You can see/download other Brainstorm models by clicking on the Brainstorm Collection on the right of your screen.