Update app.py

app.py

@@ -1,348 +1,234 @@
-# --- LangGraph & LangChain Imports ---
 import os
-import gradio as gr
 import requests
 import pandas as pd
-from typing import TypedDict, Annotated, Literal
+import gradio as gr
+from typing import Optional
 
-# --- LangGraph & LangChain Imports ---
-from langchain_core.messages import HumanMessage, AIMessage, SystemMessage
-from langgraph.graph import StateGraph, END, START
-from langgraph.graph.message import add_messages
-from langchain_groq import ChatGroq
-from langchain_core.tools import tool
-from langgraph.prebuilt import ToolNode
-from langchain_tavily import TavilySearch
-from smolagents import CodeAgent, LiteLLMModel, VisitWebpageTool, WebSearchTool
+# --- SMOLAGENTS IMPORTS ---
+from smolagents import CodeAgent, LiteLLMModel, VisitWebpageTool, DuckDuckGoSearchTool
 
-# --- Constants ---
+# --- CONSTANTS ---
 DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
 
-# --- API KEYS (Secrets se lena zaroori hai) ---
-GROQ_API_KEY = os.getenv("GROQ_API_KEY") 
-MISTRAL_API_KEY = os.getenv("MISTRAL_API_KEY")
-TAVILY_API_KEY = os.getenv("TAVILY_API_KEY")
-
+# --- AGENT CLASS (MISTRAL FOCUSED) ---
 class BasicAgent:
     def __init__(self):
-        print("Initializing LangGraph Agent (Simplified - No Critic)...")
-        
-        # --- 1. SETUP MODELS ---
-        # Main Brain (Groq)
-        self.llm = ChatGroq(
-            model="qwen/qwen3-32b", 
-            temperature=0, 
-            api_key=GROQ_API_KEY
+        print("🚀 Initializing Mistral-Powered Agent...")
+        
+        # --- 1. API KEY CHECK ---
+        mistral_key = os.getenv("MISTRAL_API_KEY")
+        if not mistral_key:
+            raise ValueError("⚠️ MISTRAL_API_KEY environment variable mein nahi mili! Settings me add karo.")
+
+        # --- 2. MODEL SETUP (Mistral Large) ---
+        # Mistral Large logic aur code ke liye best hai
+        model = LiteLLMModel(
+            model_id="mistral/mistral-large-latest",
+            api_key=mistral_key
         )
 
-        # Coding Brain (Mistral via SmolAgents)
-        model_mistral = LiteLLMModel(
-            model_id="mistral/mistral-large-latest", 
-            api_key=MISTRAL_API_KEY
+        # --- 3. TOOLS ---
+        # Web Search aur Page Visit tools
+        search_tool = DuckDuckGoSearchTool()
+        visit_tool = VisitWebpageTool()
+
+        # --- 4. CREATE AGENT ---
+        # CodeAgent use kar rahe hain jo seedha Python likhta hai
+        self.agent = CodeAgent(
+            tools=[search_tool, visit_tool],
+            model=model,
+            # GAIA ke liye ye libraries zaroori hain
+            additional_authorized_imports=[
+                "numpy", "pandas", "math", "datetime", "re", "csv", "json", "random", "itertools"
+            ],
+            max_steps=25,  # Mushkil sawalon ke liye steps badhaye
+            verbosity_level=2,
+            name="Mistral-Gaia-Solver"
         )
-        
-        # SmolAgent Tool (Complex tasks ke liye)
-        self.avii_agent = CodeAgent(
-            tools=[WebSearchTool(), VisitWebpageTool()], 
-            model=model_mistral,
-            additional_authorized_imports=["numpy", "pandas", "math", "datetime"],
-            name="avii"
-        )
-        
-        # --- 2. DEFINE TOOLS ---
-        self.web_search = TavilySearch(
-            max_results=2, 
-            include_raw_content=False, 
-            include_images=False,
-            api_key=TAVILY_API_KEY
-        )
-
-        @tool
-        def execute_code_agent(task_description: str) -> str:
-            """Use strictly for coding/calculation tasks."""
-            try:
-                sys_prompt = "You are jarvis. Solve this using Python code."
-                full_command = f"{sys_prompt}\nTask: {task_description}"
-                result = self.avii_agent.run(full_command)
-                return str(result)
-            except Exception as e:
-                return f"Error: {e}"
-
-        @tool
-        def safe_web_search(query: str) -> str:
-            """Search the web safely."""
-            try:
-                results = self.web_search.invoke(query)
-                return str(results)
-            except Exception as e:
-                return f"Search Error: {str(e)}"
 
-        tools = [execute_code_agent, safe_web_search]
-        self.llm_with_tools = self.llm.bind_tools(tools)
-        self.tool_node = ToolNode(tools)
-
-        # --- 3. BUILD GRAPH (CRITIC REMOVED) ---
-        # Simple Flow: Agent -> Router -> (Tools OR End)
-        workflow = StateGraph(self.AgentState)
-        
-        workflow.add_node("agent", self.agent_node)
-        workflow.add_node("tools", self.tool_node)
-        
-        workflow.set_entry_point("agent")
-        
-        # Router Decision
-        workflow.add_conditional_edges("agent", self.router, {
-            "tools": "tools",
-            "__end__": END
-        })
-        
-        # Tools se wapas Agent ke paas
-        workflow.add_edge("tools", "agent")
+    def __call__(self, question: str, file_path: Optional[str] = None) -> str:
+        """
+        Ye function agent ko run karta hai aur answer clean karta hai.
+        """
+        # --- PROMPT ENGINEERING FOR EXACT MATCH ---
+        prompt = f"""
+        Current Task: {question}
+        
+        INSTRUCTIONS:
+        1. You are an expert Python coding agent. Solve this step-by-step.
+        2. If a file is attached, YOU MUST READ IT using Python code immediately.
+        3. Do not assume values. Calculate them using the file or web search.
+        
+        OUTPUT FORMAT RULES:
+        - Once you find the answer, output ONLY the final value.
+        - Do not write "The answer is...".
+        - Do not add units (like $, kg, years) unless specifically asked.
+        - Example: If the answer is 42, just print 42.
+        """
         
-        self.app = workflow.compile()
-        print("Agent Graph Compiled Successfully!")
-
-    # --- Helper Classes ---
-    class AgentState(TypedDict):
-        messages: Annotated[list, add_messages]
+        if file_path:
+            prompt += f"\n\n⚠️ ATTACHED FILE: A file is available at path '{file_path}'. Read it now."
 
-    # --- Node Functions ---
-    def agent_node(self, state):
-        messages = state["messages"]
-        
-        # Context window management
-        if len(messages) > 6:
-            trimmed_messages = [messages[0]] + messages[-5:]
-        else:
-            trimmed_messages = messages
-        
-        # SYSTEM PROMPT INJECTION (Strict Formatting)
-        # Ye prompt ensure karega ki agent 'FINAL ANSWER:' format use kare
-        sys_prompt_content = (
-            "You are a general AI assistant. Solve the user's question.\n"
-            "IMPORTANT: Once you have the answer, you MUST end your response with:\n"
-            "FINAL ANSWER: [value]\n"
-            "Example: FINAL ANSWER: 2\n"
-            "Do not add extra conversational text after the final answer."
-        )
-        
-        # Ensure system prompt is always first
-        if not isinstance(trimmed_messages[0], SystemMessage):
-            trimmed_messages.insert(0, SystemMessage(content=sys_prompt_content))
+        try:
+            # Agent Run
+            print(f"🤖 Agent thinking on: {question[:50]}...")
+            response = self.agent.run(prompt)
             
-        response = self.llm_with_tools.invoke(trimmed_messages)
-        return {"messages": [response]}
-
-    def router(self, state) -> Literal["tools", "__end__"]:
-        messages = state["messages"]
-        last_message = messages[-1]
-        
-        if last_message.tool_calls:
-            return "tools"
-        return "__end__"
-
-    # --- MAIN EXECUTION (CLEANING MAGIC HERE) ---
-    def __call__(self, question: str) -> str:
-        # print(f"Agent received question: {question}")
-        
-        inputs = {"messages": [HumanMessage(content=question)]}
-        result = self.app.invoke(inputs)
-        
-        # Raw Answer from Agent
-        raw_answer = result["messages"][-1].content
-        
-        # --- CLEANING LOGIC START ---
-        # 1. 'FINAL ANSWER:' ke baad ka hissa nikalo
-        if "FINAL ANSWER:" in raw_answer:
-            final_clean = raw_answer.split("FINAL ANSWER:")[-1].strip()
-        else:
-            final_clean = raw_answer.strip()
+            # --- OUTPUT CLEANING ---
+            final_answer = str(response)
             
-        # 2. Agar end mein '.' (full stop) hai to hatao (e.g., "5." -> "5")
-        if final_clean.endswith("."):
-            final_clean = final_clean[:-1]
-        
-        # 3. Double check: Agar empty hai to original return kar do
-        if not final_clean:
-            final_clean = raw_answer
+            # Common text removal
+            final_answer = final_answer.replace("Final Answer:", "").strip()
             
-        # --- CLEANING LOGIC END ---
+            # Remove trailing dot (e.g., "100." -> "100")
+            if final_answer.endswith(".") and len(final_answer) < 20:
+                final_answer = final_answer[:-1]
+                
+            return final_answer
             
-        print(f"Original: {raw_answer[:50]}... -> Cleaned: {final_clean}")
-        return str(final_clean)
+        except Exception as e:
+            print(f"❌ Error in Agent: {e}")
+            return f"Error: {e}"
 
-def run_and_submit_all( profile: gr.OAuthProfile | None):
+# --- MAIN EVALUATION RUNNER ---
+def run_and_submit_all(profile: gr.OAuthProfile | None):
     """
-    Fetches all questions, runs the BasicAgent on them, submits all answers,
-    and displays the results.
+    1. Questions fetch karega.
+    2. FILE DOWNLOAD karega (Ye missing tha pehle).
+    3. Agent run karega.
+    4. Submit karega.
     """
-    # --- Determine HF Space Runtime URL and Repo URL ---
-    space_id = os.getenv("SPACE_ID") # Get the SPACE_ID for sending link to the code
-
-    if profile:
-        username= f"{profile.username}"
-        print(f"User logged in: {username}")
-    else:
-        print("User not logged in.")
-        return "Please Login to Hugging Face with the button.", None
-
+    
+    # --- A. LOGIN CHECK ---
+    if profile is None:
+        return "⚠️ Please Login to Hugging Face with the button above.", None
+    
+    username = profile.username
+    space_id = os.getenv("SPACE_ID")
+    
+    # URLs
     api_url = DEFAULT_API_URL
     questions_url = f"{api_url}/questions"
     submit_url = f"{api_url}/submit"
 
-    # 1. Instantiate Agent ( modify this part to create your agent)
+    # --- B. INIT AGENT ---
     try:
         agent = BasicAgent()
     except Exception as e:
-        print(f"Error instantiating agent: {e}")
-        return f"Error initializing agent: {e}", None
-    # In the case of an app running as a hugging Face space, this link points toward your codebase ( usefull for others so please keep it public)
-    agent_code = f"https://huggingface.co/spaces/{space_id}/tree/main"
-    print(agent_code)
+        return f"❌ Agent Init Error: {e}", None
 
-    # 2. Fetch Questions
-    print(f"Fetching questions from: {questions_url}")
+    agent_code_link = f"https://huggingface.co/spaces/{space_id}/tree/main"
+    print(f"🔗 Code Link: {agent_code_link}")
+
+    # --- C. FETCH QUESTIONS ---
     try:
-        response = requests.get(questions_url, timeout=15)
-        response.raise_for_status()
-        questions_data = response.json()
-        if not questions_data:
-             print("Fetched questions list is empty.")
-             return "Fetched questions list is empty or invalid format.", None
-        print(f"Fetched {len(questions_data)} questions.")
-    except requests.exceptions.RequestException as e:
-        print(f"Error fetching questions: {e}")
-        return f"Error fetching questions: {e}", None
-    except requests.exceptions.JSONDecodeError as e:
-         print(f"Error decoding JSON response from questions endpoint: {e}")
-         print(f"Response text: {response.text[:500]}")
-         return f"Error decoding server response for questions: {e}", None
+        print("📥 Fetching questions...")
+        questions_data = requests.get(questions_url).json()
     except Exception as e:
-        print(f"An unexpected error occurred fetching questions: {e}")
-        return f"An unexpected error occurred fetching questions: {e}", None
+        return f"Error fetching questions: {e}", None
 
-    # 3. Run your Agent
     results_log = []
     answers_payload = []
-    print(f"Running agent on {len(questions_data)} questions...")
+
+    print(f"🚀 Starting processing of {len(questions_data)} questions...")
+
+    # --- D. PROCESSING LOOP ---
     for item in questions_data:
-        task_id = item.get("task_id")
-        question_text = item.get("question")
-        if not task_id or question_text is None:
-            print(f"Skipping item with missing task_id or question: {item}")
-            continue
+        task_id = item["task_id"]
+        question_text = item["question"]
+        file_name = item.get("file_name") # GAIA tasks often have files
+
+        print(f"\n--- Processing Task {task_id} ---")
+        
+        local_file_path = None
+
+        # 1. DOWNLOAD FILE (CRITICAL STEP)
+        if file_name:
+            print(f"📎 Downloading file: {file_name}")
+            try:
+                file_url = f"{api_url}/files/{task_id}"
+                file_resp = requests.get(file_url, timeout=10)
+                
+                if file_resp.status_code == 200:
+                    with open(file_name, "wb") as f:
+                        f.write(file_resp.content)
+                    local_file_path = file_name
+                    print("✅ File downloaded successfully.")
+                else:
+                    print(f"❌ File download failed (Status {file_resp.status_code})")
+            except Exception as e:
+                print(f"❌ File download error: {e}")
+
+        # 2. RUN AGENT
         try:
-            submitted_answer = agent(question_text)
+            # Agent ko file path pass kar rahe hain
+            submitted_answer = agent(question_text, file_path=local_file_path)
+            
+            print(f"💡 Final Answer: {submitted_answer}")
+            
             answers_payload.append({"task_id": task_id, "submitted_answer": submitted_answer})
-            results_log.append({"Task ID": task_id, "Question": question_text, "Submitted Answer": submitted_answer})
+            results_log.append({
+                "Task ID": task_id, 
+                "Question": question_text, 
+                "File": file_name if file_name else "None",
+                "Answer": submitted_answer
+            })
+            
         except Exception as e:
-             print(f"Error running agent on task {task_id}: {e}")
-             results_log.append({"Task ID": task_id, "Question": question_text, "Submitted Answer": f"AGENT ERROR: {e}"})
+            results_log.append({"Task ID": task_id, "Error": str(e)})
 
-    if not answers_payload:
-        print("Agent did not produce any answers to submit.")
-        return "Agent did not produce any answers to submit.", pd.DataFrame(results_log)
+        # 3. CLEANUP (File delete karo)
+        if local_file_path and os.path.exists(local_file_path):
+            os.remove(local_file_path)
 
-    # 4. Prepare Submission 
-    submission_data = {"username": username.strip(), "agent_code": agent_code, "answers": answers_payload}
-    status_update = f"Agent finished. Submitting {len(answers_payload)} answers for user '{username}'..."
-    print(status_update)
+    # --- E. SUBMIT ---
+    print("📤 Submitting answers to leaderboard...")
+    submission_data = {
+        "username": username,
+        "agent_code": agent_code_link,
+        "answers": answers_payload
+    }
 
-    # 5. Submit
-    print(f"Submitting {len(answers_payload)} answers to: {submit_url}")
     try:
         response = requests.post(submit_url, json=submission_data, timeout=60)
-        response.raise_for_status()
-        result_data = response.json()
-        final_status = (
-            f"Submission Successful!\n"
-            f"User: {result_data.get('username')}\n"
-            f"Overall Score: {result_data.get('score', 'N/A')}% "
-            f"({result_data.get('correct_count', '?')}/{result_data.get('total_attempted', '?')} correct)\n"
-            f"Message: {result_data.get('message', 'No message received.')}"
+        res_json = response.json()
+        
+        score = res_json.get('score', 0)
+        correct = res_json.get('correct_count', 0)
+        
+        status_msg = (
+            f"✅ Submission Done!\n"
+            f"User: {username}\n"
+            f"🏆 Score: {score}%\n"
+            f"Correct: {correct}"
         )
-        print("Submission successful.")
-        results_df = pd.DataFrame(results_log)
-        return final_status, results_df
-    except requests.exceptions.HTTPError as e:
-        error_detail = f"Server responded with status {e.response.status_code}."
-        try:
-            error_json = e.response.json()
-            error_detail += f" Detail: {error_json.get('detail', e.response.text)}"
-        except requests.exceptions.JSONDecodeError:
-            error_detail += f" Response: {e.response.text[:500]}"
-        status_message = f"Submission Failed: {error_detail}"
-        print(status_message)
-        results_df = pd.DataFrame(results_log)
-        return status_message, results_df
-    except requests.exceptions.Timeout:
-        status_message = "Submission Failed: The request timed out."
-        print(status_message)
-        results_df = pd.DataFrame(results_log)
-        return status_message, results_df
-    except requests.exceptions.RequestException as e:
-        status_message = f"Submission Failed: Network error - {e}"
-        print(status_message)
-        results_df = pd.DataFrame(results_log)
-        return status_message, results_df
+        return status_msg, pd.DataFrame(results_log)
+
     except Exception as e:
-        status_message = f"An unexpected error occurred during submission: {e}"
-        print(status_message)
-        results_df = pd.DataFrame(results_log)
-        return status_message, results_df
+        return f"❌ Submission Failed: {e}", pd.DataFrame(results_log)
 
 
-# --- Build Gradio Interface using Blocks ---
+# --- GRADIO UI ---
 with gr.Blocks() as demo:
-    gr.Markdown("# Basic Agent Evaluation Runner")
-    gr.Markdown(
-        """
-        **Instructions:**
-
-        1.  Please clone this space, then modify the code to define your agent's logic, the tools, the necessary packages, etc ...
-        2.  Log in to your Hugging Face account using the button below. This uses your HF username for submission.
-        3.  Click 'Run Evaluation & Submit All Answers' to fetch questions, run your agent, submit answers, and see the score.
-
-        ---
-        **Disclaimers:**
-        Once clicking on the "submit button, it can take quite some time ( this is the time for the agent to go through all the questions).
-        This space provides a basic setup and is intentionally sub-optimal to encourage you to develop your own, more robust solution. For instance for the delay process of the submit button, a solution could be to cache the answers and submit in a seperate action or even to answer the questions in async.
-        """
-    )
-
+    gr.Markdown("# 🤖 GAIA Agent Solver (Mistral + Files Fix)")
+    gr.Markdown("""
+    **Instruction:**
+    1. Login via Hugging Face button.
+    2. Click 'Run Evaluation'.
+    3. Wait (it takes time to process all questions).
+    """)
+    
     gr.LoginButton()
-
-    run_button = gr.Button("Run Evaluation & Submit All Answers")
-
-    status_output = gr.Textbox(label="Run Status / Submission Result", lines=5, interactive=False)
-    # Removed max_rows=10 from DataFrame constructor
-    results_table = gr.DataFrame(label="Questions and Agent Answers", wrap=True)
-
-    run_button.click(
+    
+    run_btn = gr.Button("Run Evaluation & Submit", variant="primary")
+    
+    status_out = gr.Textbox(label="Status")
+    results_df = gr.DataFrame(label="Detailed Logs")
+    
+    run_btn.click(
         fn=run_and_submit_all,
-        outputs=[status_output, results_table]
+        outputs=[status_out, results_df]
     )
 
 if __name__ == "__main__":
-    print("\n" + "-"*30 + " App Starting " + "-"*30)
-    # Check for SPACE_HOST and SPACE_ID at startup for information
-    space_host_startup = os.getenv("SPACE_HOST")
-    space_id_startup = os.getenv("SPACE_ID") # Get SPACE_ID at startup
-
-    if space_host_startup:
-        print(f"✅ SPACE_HOST found: {space_host_startup}")
-        print(f"   Runtime URL should be: https://{space_host_startup}.hf.space")
-    else:
-        print("ℹ️  SPACE_HOST environment variable not found (running locally?).")
-
-    if space_id_startup: # Print repo URLs if SPACE_ID is found
-        print(f"✅ SPACE_ID found: {space_id_startup}")
-        print(f"   Repo URL: https://huggingface.co/spaces/{space_id_startup}")
-        print(f"   Repo Tree URL: https://huggingface.co/spaces/{space_id_startup}/tree/main")
-    else:
-        print("ℹ️  SPACE_ID environment variable not found (running locally?). Repo URL cannot be determined.")
-
-    print("-"*(60 + len(" App Starting ")) + "\n")
-
-    print("Launching Gradio Interface for Basic Agent Evaluation...")
-    demo.launch(debug=True, share=False)
+    demo.launch()