Update README.md

README.md

@@ -25,6 +25,7 @@ This model i can say ... wow ... !! ... GRPO ! It works ... I think the differen
 BUT :
 The model did not seem to be advancing , although when used as a agent it highly performs its task without the need of other models to be called : As it has been trained for each agent role:
 ## Agent Roles 
+```xml
 **Planning**
 **Coding**
 **ToolUse**
@@ -37,7 +38,7 @@ The model did not seem to be advancing , although when used as a agent it highly
 **Deep Research**
 **Deep Calculation**
 **repl**
-
+```
 Etc the list goes on !
 
 We also trainwed various scripts and personas as well as merged a few roleplay models.. Whilst Still Targetting benchmarks ! But our model improved locally but not for benchmarking !!