Spaces:
Sleeping
Sleeping
| import os | |
| import requests | |
| import gradio as gr | |
| import pandas as pd | |
| from smolagents import CodeAgent, DuckDuckGoSearchTool | |
| from smolagents.models import OpenAIServerModel | |
| from audio_transcriber import AudioTranscriptionTool | |
| from image_analyzer import ImageAnalysisTool | |
| from wikipedia_searcher import WikipediaSearcher | |
| # Static system prompt for all completions | |
| SYSTEM_PROMPT = """You are an agent solving the GAIA benchmark and you are required to provide exact answers. | |
| Rules to follow: | |
| 1. Return only the exact requested answer: no explanation and no reasoning. | |
| 2. For yes/no questions, return exactly "Yes" or "No". | |
| 3. For dates, use the exact format requested. | |
| 4. For numbers, use the exact number, no other format. | |
| 5. For names, use the exact name as found in sources. | |
| 6. If the question has an associated file, download the file first using the task ID. | |
| Examples of good responses: | |
| - "42" | |
| - "Yes" | |
| - "October 5, 2001" | |
| - "Buenos Aires" | |
| Never include phrases like "the answer is..." or "Based on my research". | |
| Only return the exact answer.""" | |
| class PatchedOpenAIServerModel(OpenAIServerModel): | |
| def generate(self, messages, stop_sequences=None, **kwargs): | |
| if isinstance(messages, list): | |
| if not any(m["role"] == "system" for m in messages): | |
| messages = [{"role": "system", "content": SYSTEM_PROMPT}] + messages | |
| return super().generate(messages=messages, stop_sequences=stop_sequences, **kwargs) | |
| class MyAgent: | |
| def __init__(self): | |
| self.model = PatchedOpenAIServerModel(model_id="gpt-4-turbo") | |
| self.agent = CodeAgent( | |
| tools=[ | |
| DuckDuckGoSearchTool(), | |
| WikipediaSearcher(), | |
| AudioTranscriptionTool(), | |
| ImageAnalysisTool() | |
| ], | |
| model=self.model | |
| ) | |
| def __call__(self, task: dict) -> str: | |
| question = task.get("question", "") | |
| attachment = task.get("attachment") | |
| if attachment: | |
| if attachment.endswith((".mp3", ".wav")): | |
| question += f"\n\nAudio file: {attachment}" | |
| elif attachment.endswith((".jpg", ".jpeg", ".png")): | |
| question += f"\n\nImage file: {attachment}" | |
| elif attachment.endswith(".py"): | |
| try: | |
| content = requests.get(attachment).text | |
| question += f"\n\nPython code:\n{content}" | |
| except Exception as e: | |
| question += f"\n\nError loading code: {e}" | |
| return self.agent.run(question) | |
| def run_and_submit_all(profile: gr.OAuthProfile | None): | |
| space_id = os.getenv("SPACE_ID") | |
| if profile: | |
| username = 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" | |
| try: | |
| agent = MyAgent() | |
| except Exception as e: | |
| print(f"Error initializing agent: {e}") | |
| return f"Error initializing agent: {e}", None | |
| agent_code = f"https://huggingface.co/spaces/{space_id}/tree/main" | |
| print(f"Agent code URL: {agent_code}") | |
| 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: | |
| return "Fetched questions list is empty or invalid format.", None | |
| print(f"Fetched {len(questions_data)} questions.") | |
| except Exception as e: | |
| return f"Error fetching questions: {e}", None | |
| results_log = [] | |
| answers_payload = [] | |
| print(f"Running agent on {len(questions_data)} questions...") | |
| for item in questions_data: | |
| task_id = item.get("task_id") | |
| if not task_id: | |
| continue | |
| try: | |
| submitted_answer = agent(item) | |
| answers_payload.append({"task_id": task_id, "submitted_answer": submitted_answer}) | |
| results_log.append({ | |
| "Task ID": task_id, | |
| "Question": item.get("question", ""), | |
| "Submitted Answer": submitted_answer | |
| }) | |
| except Exception as e: | |
| error_msg = f"AGENT ERROR: {e}" | |
| results_log.append({ | |
| "Task ID": task_id, | |
| "Question": item.get("question", ""), | |
| "Submitted Answer": error_msg | |
| }) | |
| if not answers_payload: | |
| return "Agent did not produce any answers to submit.", pd.DataFrame(results_log) | |
| submission_data = { | |
| "username": username.strip(), | |
| "agent_code": agent_code, | |
| "answers": answers_payload | |
| } | |
| 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.')}" | |
| ) | |
| results_df = pd.DataFrame(results_log) | |
| return final_status, results_df | |
| except requests.exceptions.HTTPError as e: | |
| try: | |
| detail = e.response.json().get("detail", e.response.text) | |
| except Exception: | |
| detail = e.response.text[:500] | |
| return f"Submission Failed: {detail}", pd.DataFrame(results_log) | |
| except requests.exceptions.Timeout: | |
| return "Submission Failed: The request timed out.", pd.DataFrame(results_log) | |
| except Exception as e: | |
| return f"An unexpected error occurred during submission: {e}", pd.DataFrame(results_log) | |
| # Gradio UI setup | |
| with gr.Blocks() as demo: | |
| gr.Markdown("# Basic Agent Evaluation Runner") | |
| gr.Markdown(""" | |
| **Instructions:** | |
| 1. Clone this space, modify code to define your agent's logic, tools, and packages. | |
| 2. Log in to your Hugging Face account using the button below. | |
| 3. Click 'Run Evaluation & Submit All Answers' to fetch questions, run your agent, submit answers, and see your score. | |
| **Note:** Submitting can take some time. | |
| """) | |
| 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) | |
| space_host = os.getenv("SPACE_HOST") | |
| space_id = os.getenv("SPACE_ID") | |
| if space_host: | |
| print(f"✅ SPACE_HOST found: {space_host}") | |
| print(f" Runtime URL should be: https://{space_host}.hf.space") | |
| else: | |
| print("ℹ️ SPACE_HOST environment variable not found (running locally?).") | |
| if space_id: | |
| print(f"✅ SPACE_ID found: {space_id}") | |
| print(f" Repo URL: https://huggingface.co/spaces/{space_id}") | |
| print(f" Repo Tree URL: https://huggingface.co/spaces/{space_id}/tree/main") | |
| else: | |
| print("ℹ️ SPACE_ID environment variable not found (running locally?).") | |
| print("-"*(60 + len(" App Starting ")) + "\n") | |
| print("Launching Gradio Interface for Basic Agent Evaluation...") | |
| demo.launch(debug=True, share=False) | |