Update README.md

README.md

@@ -31,10 +31,10 @@ The Multi-Agentic SQL Generator is a modular, multi-step system that translates
 
 - **Evaluation with RAGAS Metrics:**  
   Evaluates the generated output using metrics like:
-  - **Faithfulness:** How closely the output matches the expected result.
-  - **Answer Relevancy:** How well the result addresses the user query.
-  - **Context Precision:** Whether the query returns only the necessary data.
-  - **Context Recall:** Whether the query returns all required data.
+  - **Faithfulness:** (0.7500) Faithfulness measures how well the generated SQL execution results match the expected (ground truth) results.
+  - **Answer Relevancy:** (0.2280) Answer relevancy measures how well the generated SQL results match the intent of the user's query.
+  - **Context Precision:** (0.3000) Context Precision measures how much of the generated SQL execution result is actually relevant to the query.
+  - **Context Recall:** (0.3500) Context Recall measures how much of the expected result was actually retrieved.
 
 - **Extensibility and Deployment:**  
   Easily integrable with front-end frameworks (e.g., Chainlit) and deployable on platforms like Hugging Face Spaces.