Update app.py

app.py

@@ -20,96 +20,18 @@ MODEL_ID = "TheBloke/Mistral-7B-Instruct-v0.1-GPTQ"  # 4.2GB quantized
 EMBEDDING_MODEL = "sentence-transformers/all-mpnet-base-v2"
 FALLBACK_MODELS = ["google/flan-t5-base", "mistralai/Mistral-7B-Instruct-v0.2"]
 
-class AgentState(TypedDict):
-    question: str
-    context: str
-    answer: str
-    attempts: Annotated[int, lambda x, y: x + 1]
-
+# --- Basic Agent Definition ---
+# ----- THIS IS WERE YOU CAN BUILD WHAT YOU WANT ------
 class BasicAgent:
+    
     def __init__(self):
-        # Initialize components
-        self.client = InferenceClient(
-            model=MODEL_ID,
-            token=os.environ["HF_TOKEN"],
-            timeout=120
-        )
-        
-        # Initialize vector store (add your documents here)
-        self.vectorstore = Chroma.from_texts(
-            texts=["GAIA knowledge content..."],  # Replace with your documents
-            embedding=EMBEDDING_MODEL,
-            persist_directory="./chroma_db"
-        )
+        print("BasicAgent initialized.")
         
-        # Build LangGraph workflow
-        self.workflow = self._build_graph()
-
-    def _build_graph(self):
-        # Define nodes
-        def retrieve(state: AgentState):
-            docs = self.vectorstore.similarity_search(state["question"], k=3)
-            state["context"] = "\n".join([d.page_content for d in docs])
-            return state
-
-        def generate(state: AgentState):
-            try:
-                response = self.client.text_generation(
-                    f"""<s>[INST]Answer using ONLY this context:
-                    {state['context']}
-                    Question: {state['question']}
-                    Answer: [/INST]""",
-                    temperature=0.1,
-                    max_new_tokens=100,
-                    stop_sequences=["</s>"]
-                )
-                state["answer"] = response.split("[/INST]")[-1].strip()
-            except Exception:
-                state["answer"] = ""
-            return state
-
-        def validate(state: AgentState):
-            if len(state["answer"]) > 5 and state["attempts"] < 3:
-                return "final"
-            return "retry"
-
-        # Build workflow
-        workflow = StateGraph(AgentState)
-        workflow.add_node("retrieve", retrieve)
-        workflow.add_node("generate", generate)
-        workflow.add_node("validate", validate)
-
-        workflow.set_entry_point("retrieve")
-        workflow.add_edge("retrieve", "generate")
-        workflow.add_edge("generate", "validate")
-        
-        workflow.add_conditional_edges(
-            "validate",
-            lambda x: "retry" if x["answer"] == "" else "final",
-            {
-                "retry": "retrieve",
-                "final": END
-            }
-        )
-
-        return workflow.compile()
-
     def __call__(self, question: str) -> str:
-        # GAIA-compliant formatting
-        state = {
-            "question": question,
-            "context": "",
-            "answer": "",
-            "attempts": 0
-        }
-        
-        for _ in range(3):  # Max 3 attempts
-            state = self.workflow.invoke(state)
-            if state["answer"]:
-                answer = re.sub(r'[^a-zA-Z0-9]', '', state["answer"]).lower()
-                return answer[:100]  # GAIA length constraint
-        
-        return ""  # Preserve scoring eligibility
+        print(f"Agent received question (first 50 chars): {question[:50]}...")
+        fixed_answer = "This is a default answer."
+        print(f"Agent returning fixed answer: {fixed_answer}")
+        return fixed_answer
         
 def run_and_submit_all( profile: gr.OAuthProfile | None):
     """