Update main.py

main.py

@@ -11,10 +11,9 @@ llm = Llama(
 
 # Pydantic object for validation
 class Validation(BaseModel):
-    user_prompt: str
-    system_prompt: str
-    max_tokens: int = 1024  
-    temperature: float = 0.01  
+    user_prompt: str  # This will be the direct SQL query request or relevant prompt
+    max_tokens: int = 1024
+    temperature: float = 0.01
 
 # FastAPI application initialization
 app = FastAPI()
@@ -22,11 +21,8 @@ app = FastAPI()
 # Endpoint for generating responses
 @app.post("/generate_response")
 async def generate_response(item: Validation):
-    # Construct the complete prompt using the given system and user prompts in the required format
-    prompt = f"<|user|>\n{item.system_prompt}\n {item.user_prompt}\n<|end|>\n<|assistant|>"
-
-    # Call the Llama model to generate a response
-    output = llm(prompt, max_tokens=item.max_tokens, temperature=item.temperature, echo=True)
+    # Call the Llama model to generate a response directly based on the user's prompt
+    output = llm(item.user_prompt, max_tokens=item.max_tokens, temperature=item.temperature, echo=False)
     
     # Extract and return the text from the response
     return output['choices'][0]['text']