Upload 2 files

src/agent.py

@@ -23,7 +23,7 @@ class Agent:
         self.temperature = temperature  # Default value
         self.top_p = top_p  # Default value
 
-    def process_query(self, user_id: str, query: str, context: str = "") -> str:
+    def process_query(self, user_id: str, query: str) -> str:
         # Normalize the query to lowercase
         query = query.lower()
 
@@ -53,7 +53,7 @@ class Agent:
                 return response
 
         # For general queries, use the existing multi-layer processing
-        initial_response = extract_and_summarize(query, self.memory, self.llm, self.dynamic_query_response_prompt(query), max_tokens=self.max_tokens, temperature=self.temperature, top_p=self.top_p, context=context)
+        initial_response = extract_and_summarize(query, self.memory, self.llm, self.dynamic_query_response_prompt(query), max_tokens=self.max_tokens, temperature=self.temperature, top_p=self.top_p)
 
         # Evaluate the initial response
         if not self.evaluate_response(initial_response, query):
@@ -62,7 +62,7 @@ class Agent:
             # Combine initial and additional data
             combined_context = f"{initial_response}\n{additional_data}"
             # Truncate the combined context to fit within the model's context window
-            max_context_length = 500  # Adjust this based on your LLM's token limit
+            max_context_length = 30000  # Adjust this based on your LLM's token limit
             if len(combined_context) > max_context_length:
                 combined_context = combined_context[:max_context_length]
                 logging.info(f"Truncated combined context to {max_context_length} characters.")