Update README.md

README.md

@@ -19,7 +19,7 @@ tags:
 ## Description:
 
 **PENDING**  
-A Starling-based model focused on being an assistant worth talking to.  
+A companion focused on being an assistant worth talking to. Now better at reasoning!
 
 **Uncensored**
 
@@ -35,8 +35,6 @@ Velara was designed to address specific issues found in other chat models:
 
 - **STRICTLY Sticking to the Prompt/Prompt Fidelity:** More care was made to make sure she pays attention to the pretext and overall conversation. This is mostly shown and used for the "addons" section, but it applies to everything in the prompt.
 
-- **Addons, Sort Of:** Something I had tried in the past and wanted to add to a model deliberately was the ability to give it "addons". You can add any amount of "addons" to Velara's mind. These aren't directly controllable. You can then ask her what they do, or if you have the LTS installed, ask her and change the reply to what you'd want it to do.
-
 ## Prompt Template:
 
 **For optimal interaction, use this template:**
@@ -57,21 +55,10 @@ Always take the entire conversation into account when forming and writing a repl
 **Defaults:**
 
 ```
-temperature: 0.8
-top_p: 0.85
-top_k: 20
-repetition_penalty: 1.2
-guidance_scale: 1.25
-```
-
-
-**Worse context understanding but less repetitive in some cases:**
-
-```
-temperature: 0.77
-top_p: 0.85
-top_k: 20
-repetition_penalty: 1.2
+min_p: 0.2
+repetition_penalty: 1.13
+repetition_penalty_range: 0
+guidance_scale: 1.05
 ```
 
 # Benchmarks: