Update app.py

app.py

@@ -2,191 +2,65 @@ import os
 import gradio as gr
 import requests
 import pandas as pd
-from groq import Groq
+from io import BytesIO
 
-# --- New Imports for LangChain Agent ---
+# --- LangChain & Groq Imports ---
+from groq import Groq
 from langchain_groq import ChatGroq
 from langchain.agents import AgentExecutor, create_tool_calling_agent
 from langchain_community.tools.tavily_search import TavilySearchResults
 from langchain_core.prompts import ChatPromptTemplate
+from langchain.tools import Tool
 
 
 # --- Constants ---
 DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
 
 
-# --- Agent Definition ---
-# This new agent uses LangChain to orchestrate an LLM with tools.
-class LangChainAgent:
-    def __init__(self, groq_api_key, tavily_api_key):
-        """
-        Initializes the agent with an LLM and a set of tools.
-        """
-        print("Initializing LangChainAgent...")
-
-        # 1. Initialize the LLM
-        # We use ChatGroq, the LangChain integration for Groq's API.
-        self.llm = ChatGroq(
-            model_name="llama3-70b-8192",
-            groq_api_key=groq_api_key,
-            temperature=0.0
-        )
-
-        # 2. Define the tools the agent can use
-        # For now, we'll just give it a web search tool.
-        self.tools = [
-            TavilySearchResults(max_results=3, tavily_api_key=tavily_api_key)
-        ]
-
-        # 3. Create the Agent Prompt
-        # This tells the agent how to behave and how to use the tools.
-        prompt = ChatPromptTemplate.from_messages(
-            [
-                ("system", "You are a helpful assistant. You have access to a web search tool. Respond with the final answer to the user's question."),
-                ("placeholder", "{chat_history}"),
-                ("human", "{input}"),
-                ("placeholder", "{agent_scratchpad}"),
-            ]
-        )
-
-        # 4. Create the Agent itself
-        agent = create_tool_calling_agent(self.llm, self.tools, prompt)
-
-        # 5. Create the Agent Executor
-        # This is the runtime that will actually execute the agent's logic.
-        self.agent_executor = AgentExecutor(
-            agent=agent,
-            tools=self.tools,
-            verbose=True # Set to True to see the agent's thought process
-        )
-        print("LangChainAgent initialized.")
-
-
-    def __call__(self, question: str) -> str:
-        """
-        This method is called to answer a question.
-        It invokes the agent executor.
-        """
-        print(f"LangChainAgent received question (first 50 chars): {question[:50]}...")
-        
-        # We need to handle the case where the agent makes a mistake
-        try:
-            response = self.agent_executor.invoke({"input": question})
-            answer = response.get("output", "No answer found.")
-        except Exception as e:
-            print(f"An error occurred in the agent executor: {e}")
-            answer = f"Agent failed with an error: {e}"
-
-        print(f"LangChainAgent generated answer: {answer}")
-        return answer
-
-
-def run_and_submit_all(profile: gr.OAuthProfile | None):
+# --- Custom Tool Definition using Groq ---
+def transcribe_audio_from_task_id(task_id: str) -> str:
     """
-    Fetches questions, runs the LangChainAgent on them, submits the answers,
-    and displays the results.
+    Downloads an audio file for a given task_id from the scoring server,
+    transcribes it using the GROQ API with Whisper, and returns the text.
+    Use this tool ONLY when a question explicitly mentions an audio file or recording.
+    The task_id MUST be provided as the input.
     """
-    # --- Authentication and Setup ---
-    space_id = os.getenv("SPACE_ID")
-    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
-
-    api_url = DEFAULT_API_URL
-    questions_url = f"{api_url}/questions"
-    submit_url = f"{api_url}/submit"
-
-    # 1. Instantiate Agent (using the new LangChainAgent)
+    print(f"Tool 'transcribe_audio_from_task_id' (using Groq) called with task_id: {task_id}")
     try:
-        groq_api_key = os.getenv("GROQ_API_KEY")
-        tavily_api_key = os.getenv("TAVILY_API_KEY")
-        if not groq_api_key or not tavily_api_key:
-            raise ValueError("API Keys (GROQ_API_KEY, TAVILY_API_KEY) not found in secrets.")
-        agent = LangChainAgent(groq_api_key=groq_api_key, tavily_api_key=tavily_api_key)
-    except Exception as e:
-        print(f"Error instantiating agent: {e}")
-        return f"Error initializing agent: {e}", None
-
-    agent_code = f"https://huggingface.co/spaces/{space_id}/tree/main"
-    print(f"Agent code link: {agent_code}")
-
-    # 2. Fetch Questions (same as before)
-    print(f"Fetching questions from: {questions_url}")
-    try:
-        response = requests.get(questions_url, timeout=20)
-        response.raise_for_status()
-        questions_data = response.json()
-    except Exception as e:
-        return f"Error fetching questions: {e}", None
-
-    # 3. Run your Agent (same as before)
-    results_log = []
-    answers_payload = []
-    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:
-            continue
-        submitted_answer = agent(question_text)
-        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})
-
-    # 4. Prepare Submission (same as before)
-    submission_data = {"username": username.strip(), "agent_code": agent_code, "answers": answers_payload}
-
-    # 5. Submit (same as before)
-    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.')}"
+        # Step 1: Download the file
+        file_url = f"{DEFAULT_API_URL}/files/{task_id}"
+        print(f"Downloading audio file from: {file_url}")
+        audio_response = requests.get(file_url)
+        audio_response.raise_for_status()
+
+        # Step 2: Prepare the file for the Groq API
+        # The API expects a file-like object with a name.
+        audio_bytes = BytesIO(audio_response.content)
+        audio_bytes.name = f"{task_id}.mp3" # Give the file-like object a name
+
+        # Step 3: Initialize the Groq client and transcribe
+        print("Initializing Groq client for transcription...")
+        client = Groq(api_key=os.getenv("GROQ_API_KEY"))
+
+        print("Transcribing audio with Groq's Whisper...")
+        transcription = client.audio.transcriptions.create(
+          file=audio_bytes,
+          model="whisper-large-v3",
+          response_format="text",
         )
-        results_df = pd.DataFrame(results_log)
-        return final_status, results_df
+        
+        transcribed_text = str(transcription)
+        print(f"Transcription successful. Result: {transcribed_text}")
+        return transcribed_text
+
     except Exception as e:
-        status_message = f"An unexpected error occurred during submission: {e}"
-        results_df = pd.DataFrame(results_log)
-        return status_message, results_df
+        error_message = f"Error in Groq audio transcription tool: {e}"
+        print(error_message)
+        return error_message
 
-# --- Build Gradio Interface (Mostly the same) ---
-with gr.Blocks() as demo:
-    gr.Markdown("# LangChain Agent Evaluation Runner")
-    gr.Markdown(
-        """
-        **Instructions:**
-        1.  Make sure you have set `GROQ_API_KEY` and `TAVILY_API_KEY` in your Space's secrets.
-        2.  Log in below. This is required for submission.
-        3.  Click 'Run Evaluation' to start the agent. You can see its thought process in the application logs!
-        """
-    )
-    gr.LoginButton()
-    run_button = gr.Button("Run Evaluation & Submit All Answers")
-    status_output = gr.Textbox(label="Run Status / Submission Result", lines=5, interactive=False)
-    results_table = gr.DataFrame(label="Questions and Agent Answers", wrap=True)
-    run_button.click(
-        fn=run_and_submit_all,
-        outputs=[status_output, results_table]
-    )
 
-if __name__ == "__main__":
-    print("\n" + "-"*30 + " App Starting " + "-"*30)
-    # Startup checks for secrets
-    if not os.getenv("GROQ_API_KEY"):
-        print("⚠️ WARNING: GROQ_API_KEY secret not set.")
-    else:
-        print("✅ GROQ_API_KEY secret is set.")
-    if not os.getenv("TAVILY_API_KEY"):
-        print("⚠️ WARNING: TAVILY_API_KEY secret not set.")
-    else:
-        print("✅ TAVILY_API_KEY secret is set.")
-    print("-"*(60 + len(" App Starting ")) + "\n")
-    demo.launch(debug=True, share=False)
+# --- Agent Definition ---
+class LangChainAgent:
+    def __init__(self, groq_api_key: str, tavily_api_key: str):
+        print("Initializing LangChainAgent...")
+        self.llm = ChatGroq(model_name="llama3-70b-8192", groq_api_key=groq