Update app.py

app.py

@@ -1,197 +1,266 @@
+# --- Basic Agent Definition ---
+import asyncio
 import os
-import gradio as gr
-import requests
-import inspect
+import sys
+import logging
+import random
 import pandas as pd
-from agent import BasicAgent
+import requests
+import wikipedia as wiki
+from markdownify import markdownify as to_markdown
+from typing import Any
+from dotenv import load_dotenv
+from google.generativeai import types, configure
 
-# (Keep Constants as is)
-# --- Constants ---
-DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
+from smolagents import InferenceClientModel, LiteLLMModel, ToolCallingAgent, Tool, DuckDuckGoSearchTool
 
-# --- Basic Agent Definition ---
-# ----- THIS IS WERE YOU CAN BUILD WHAT YOU WANT ------
-#class BasicAgent:
-#   def __init__(self):
-#        print("BasicAgent initialized.")
-#    def __call__(self, question: str) -> str:
-#        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):
-    """
-    Fetches all questions, runs the BasicAgent on them, submits all answers,
-    and displays the results.
-    """
-    # --- Determine HF Space Runtime URL and Repo URL ---
-    space_id = os.getenv("SPACE_ID") # Get the SPACE_ID for sending link to the code
+# Load environment and configure Gemini
+load_dotenv()
+configure(api_key=os.getenv("GOOGLE_API_KEY"))
 
-    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 ( modify this part to create your 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)
-
-    # 2. Fetch Questions
-    print(f"Fetching questions from: {questions_url}")
-    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
-    except Exception as e:
-        print(f"An unexpected error occurred fetching questions: {e}")
-        return f"An unexpected error occurred fetching questions: {e}", None
-
-    # 3. Run your Agent
-    results_log = []
-    answers_payload = []
-    print(f"Running agent on {len(questions_data)} questions...")
-    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
+# Logging
+#logging.basicConfig(level=logging.INFO, format="%(asctime)s | %(levelname)s | %(message)s")
+#logger = logging.getLogger(__name__)
+
+# --- Model Configuration ---
+GEMINI_MODEL_NAME = "gemini/gemini-2.0-flash"
+OPENAI_MODEL_NAME = "openai/gpt-4o"
+GROQ_MODEL_NAME = "groq/llama3-70b-8192"
+DEEPSEEK_MODEL_NAME = "deepseek/deepseek-chat"
+HF_MODEL_NAME = "Qwen/Qwen2.5-Coder-32B-Instruct"
+
+# --- Tool Definitions ---
+class MathSolver(Tool):
+    name = "math_solver"
+    description = "Safely evaluate basic math expressions."
+    inputs = {"input": {"type": "string", "description": "Math expression to evaluate."}}
+    output_type = "string"
+
+    def forward(self, input: str) -> str:
         try:
-            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})
+            return str(eval(input, {"__builtins__": {}}))
         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}"})
-
-    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)
-
-    # 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)
-
-    # 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.')}"
-        )
-        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}."
+            return f"Math error: {e}"
+
+class RiddleSolver(Tool):
+    name = "riddle_solver"
+    description = "Solve basic riddles using logic."
+    inputs = {"input": {"type": "string", "description": "Riddle prompt."}}
+    output_type = "string"
+
+    def forward(self, input: str) -> str:
+        if "forward" in input and "backward" in input:
+            return "A palindrome"
+        return "RiddleSolver failed."
+
+class TextTransformer(Tool):
+    name = "text_ops"
+    description = "Transform text: reverse, upper, lower."
+    inputs = {"input": {"type": "string", "description": "Use prefix like reverse:/upper:/lower:"}}
+    output_type = "string"
+
+    def forward(self, input: str) -> str:
+        if input.startswith("reverse:"):
+            reversed_text = input[8:].strip()[::-1]
+            if 'left' in reversed_text.lower():
+                return "right"
+            return reversed_text
+        if input.startswith("upper:"):
+            return input[6:].strip().upper()
+        if input.startswith("lower:"):
+            return input[6:].strip().lower()
+        return "Unknown transformation."
+
+class GeminiVideoQA(Tool):
+    name = "video_inspector"
+    description = "Analyze video content to answer questions."
+    inputs = {
+        "video_url": {"type": "string", "description": "URL of video."},
+        "user_query": {"type": "string", "description": "Question about video."}
+    }
+    output_type = "string"
+
+    def __init__(self, model_name, *args, **kwargs):
+        super().__init__(*args, **kwargs)
+        self.model_name = model_name
+
+    def forward(self, video_url: str, user_query: str) -> str:
+        req = {
+            'model': f'models/{self.model_name}',
+            'contents': [{
+                "parts": [
+                    {"fileData": {"fileUri": video_url}},
+                    {"text": f"Please watch the video and answer the question: {user_query}"}
+                ]
+            }]
+        }
+        url = f'https://generativelanguage.googleapis.com/v1beta/models/{self.model_name}:generateContent?key={os.getenv("GOOGLE_API_KEY")}'
+        res = requests.post(url, json=req, headers={'Content-Type': 'application/json'})
+        if res.status_code != 200:
+            return f"Video error {res.status_code}: {res.text}"
+        parts = res.json()['candidates'][0]['content']['parts']
+        return "".join([p.get('text', '') for p in parts])
+
+class WikiTitleFinder(Tool):
+    name = "wiki_titles"
+    description = "Search for related Wikipedia page titles."
+    inputs = {"query": {"type": "string", "description": "Search query."}}
+    output_type = "string"
+
+    def forward(self, query: str) -> str:
+        results = wiki.search(query)
+        return ", ".join(results) if results else "No results."
+
+class WikiContentFetcher(Tool):
+    name = "wiki_page"
+    description = "Fetch Wikipedia page content."
+    inputs = {"page_title": {"type": "string", "description": "Wikipedia page title."}}
+    output_type = "string"
+
+    def forward(self, page_title: str) -> str:
         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
-    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
-
-
-# --- Build Gradio Interface using Blocks ---
-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.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(
-        fn=run_and_submit_all,
-        outputs=[status_output, results_table]
-    )
+            return to_markdown(wiki.page(page_title).html())
+        except wiki.exceptions.PageError:
+            return f"'{page_title}' not found."
 
-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?).")
+class FileAttachmentQueryTool(Tool):
+    name = "run_query_with_file"
+    description = """
+    Downloads a file mentioned in a user prompt, adds it to the context, and runs a query on it.
+    This assumes the file is 20MB or less.
+    """
+    inputs = {
+        "task_id": {
+            "type": "string",
+            "description": "A unique identifier for the task related to this file, used to download it."
+        },
+        "mime_type": {
+            "type": "string",
+            "nullable": True,
+            "description": "The MIME type of the file, or the best guess if unknown."
+        },
+        "user_query": {
+            "type": "string",
+            "description": "The question to answer about the file."
+        }
+    }
+    output_type = "string"
 
-    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.")
+    def forward(self, task_id: str, mime_type: str | None, user_query: str) -> str:
+        file_url = f"https://agents-course-unit4-scoring.hf.space/files/{task_id}"
+        file_response = requests.get(file_url)
+        if file_response.status_code != 200:
+            return f"Failed to download file: {file_response.status_code} - {file_response.text}"
+        file_data = file_response.content
+        mime_type = mime_type or file_response.headers.get('Content-Type', 'application/octet-stream')
+
+        from google.generativeai import GenerativeModel
+        model = GenerativeModel(self.model_name)
+        response = model.generate_content([
+            types.Part.from_bytes(data=file_data, mime_type=mime_type),
+            user_query
+        ])
+
+        return response.text
+
+# --- Basic Agent Definition ---
+class BasicAgent:
+    def __init__(self, provider="deepseek"):
+        print("BasicAgent initialized.")
+        model = self.select_model(provider)
+        client = InferenceClientModel()
+        tools = [
+            DuckDuckGoSearchTool(),
+            GeminiVideoQA(GEMINI_MODEL_NAME),
+            WikiTitleFinder(),
+            WikiContentFetcher(),
+            MathSolver(),
+            RiddleSolver(),
+            TextTransformer(),
+            FileAttachmentQueryTool(model_name=GEMINI_MODEL_NAME),
+        ]
+        self.agent = ToolCallingAgent(
+            model=model,
+            tools=tools,
+            add_base_tools=False,
+            max_steps=10,
+        )
+        self.agent.system_prompt = (
+            """
+            You are a GAIA benchmark AI assistant. Your sole purpose is to provide exact, minimal answers in the format 'FINAL ANSWER: [ANSWER]' with no additional text, explanations, or comments. 
 
-    print("-"*(60 + len(" App Starting ")) + "\n")
+            - If the answer is a number, use numerals (e.g., '42', not 'forty-two'), without commas or units (e.g., no '$', '%') unless explicitly requested.
+            - If the answer is a string, use no articles ('a', 'the'), no abbreviations (e.g., 'New York', not 'NY'), and write digits as text (e.g., 'one', not '1') unless specified.
+            - For comma-separated lists, apply the above rules to each element based on whether it's a number or string.
+            - Answer as literally as possible, making minimal assumptions and adhering to the question's narrowest interpretation.
+            - For videos, analyze the entire content but extract only the precise answer to the query, ignoring irrelevant details.
+            - For Wikipedia or search tools, distill results to the minimal correct answer, ignoring extraneous content.
+            - If proving something, compute step-by-step internally but output only the final result in the required format.
+            - If tool outputs are verbose, extract only the essential answer that satisfies the question.
+            - Under no circumstances include explanations, intermediate steps, or text outside the 'FINAL ANSWER: [ANSWER]' format.
 
-    print("Launching Gradio Interface for Basic Agent Evaluation...")
-    demo.launch(debug=True, share=False)
+            Example:
+            Question: What is 2 + 2?
+            Response: FINAL ANSWER: 4
+
+            Your response must always be:
+            FINAL ANSWER: [ANSWER]
+            """
+        )
+
+    def select_model(self, provider: str):
+        if provider == "openai":
+            return LiteLLMModel(model_id=OPENAI_MODEL_NAME, api_key=os.getenv("OPENAI_API_KEY"))
+        elif provider == "groq":
+            return LiteLLMModel(model_id=GROQ_MODEL_NAME, api_key=os.getenv("GROQ_API_KEY"))
+        elif provider == "deepseek":
+            return LiteLLMModel(model_id=DEEPSEEK_MODEL_NAME, api_key=os.getenv("DEEPSEEK_API_KEY"))
+        elif provider == "hf":
+            return InferenceClientModel()
+        else:
+            return LiteLLMModel(model_id=GEMINI_MODEL_NAME, api_key=os.getenv("GOOGLE_API_KEY"))
+
+    def __call__(self, question: str) -> str:
+        print(f"Agent received question (first 50 chars): {question[:50]}...")
+        result = self.agent.run(question)
+        if isinstance(result, dict) and "final_answer" in result and isinstance(result["final_answer"], str):
+            final_str = result["final_answer"].strip()
+        else:
+            final_str = str(result).strip()
+
+        return f"FINAL ANSWER: {final_str}"
+
+    def evaluate_random_questions(self, csv_path: str = "gaia_qa.csv", sample_size: int = 3, show_steps: bool = True):
+        df = pd.read_csv(csv_path)
+        if not {"question", "answer"}.issubset(df.columns):
+            print("CSV must contain 'question' and 'answer' columns.")
+            print("Found columns:", df.columns.tolist())
+            return
+        samples = df.sample(n=sample_size)
+        for _, row in samples.iterrows():
+            question = row["question"].strip()
+            expected = f"FINAL ANSWER: {str(row['answer']).strip()}"
+            result = self(question).strip()
+            if show_steps:
+                print("---")
+                print("Question:", question)
+                print("Expected:", expected)
+                print("Agent:", result)
+                print("Correct:", expected == result)
+            else:
+                print(f"Q: {question}\nE: {expected}\nA: {result}\n✓: {expected == result}\n")
+
+if __name__ == "__main__":
+    args = sys.argv[1:]
+    if not args or args[0] in {"-h", "--help"}:
+        print("Usage: python agent.py [question | dev]")
+        print(" - Provide a question to get a GAIA-style answer.")
+        print(" - Use 'dev' to evaluate 3 random GAIA questions from gaia_qa.csv.")
+        sys.exit(0)
+
+    q = " ".join(args)
+    agent = BasicAgent()
+    if q == "dev":
+        agent.evaluate_random_questions()
+    else:
+        print(agent(q))