Update README.md

README.md

@@ -16,23 +16,10 @@ metrics:
 **Model Type:** Roleplay and Conversational AI  
 **Architecture:** Llama-3.1
 
-NarrativAI Llama 3.1 8B is a large-scale open-source language model built with a focus on enhancing roleplay and interactive storytelling experiences. It uses a self-improving training method called **Reflection-Tuning**, which enables the model to learn from its own outputs to reduce errors over time. While not perfect, NarrativAI is designed to be adaptive, continuously refining its narrative capabilities for deeper and more engaging interactions in roleplaying contexts.
-
-## Key Features
-
-- **Reflection-Tuning:** This training technique allows the model to evaluate its own outputs, recognizing and learning from mistakes. Over time, this process improves the model's ability to generate more accurate and contextually appropriate responses.
-- **Roleplay Focus:** Designed specifically for roleplaying and storytelling, NarrativAI can craft dynamic characters, plot twists, and interactions based on user input, all while attempting to keep responses coherent and engaging.
-- **Continuous Learning:** The model is not immune to occasional inaccuracies or hallucinations, but through Reflection-Tuning, it is actively learning to minimize such instances by self-assessing and improving with each iteration.
-
-## Training Details
-
-- **Training Technique:** NarrativAI was trained using **Reflection-Tuning**, a method where the model's responses are analyzed in feedback loops. This allows the AI to identify where it may have misunderstood or gone off course, enabling gradual improvements in accuracy and narrative consistency.
-- **Data and Scope:** The training data includes diverse narrative and conversational contexts to help the model generate rich roleplay scenarios across a variety of genres and settings.
-  
 
 ## Ethical Considerations
 
-- **Hallucinations:** Like most language models, NarrativAI can sometimes generate outputs that are factually incorrect or contextually inconsistent. Its Reflection-Tuning approach aims to reduce these occurrences over time, though users should remain mindful that the model is still evolving in this respect.
+- **Hallucinations:** Like most language models, NarrativAI can sometimes generate outputs that are factually incorrect or contextually inconsistent. 
 - **Transparency:** As an open-source model, NarrativAI welcomes community feedback to continue improving its reliability and creative potential.
 
 ## Getting Started