""" Basic Agent Evaluation Runner""" import os import inspect import gradio as gr import requests import pandas as pd import time from langchain_core.messages import HumanMessage from agent import build_graph import re # (Keep Constants as is) # --- Constants --- DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space" # --- Basic Agent Definition --- # ----- THIS IS WERE YOU CAN BUILD WHAT YOU WANT ------ cached_answers = [] def parse_final_answer(raw: str) -> str: raw = raw.strip() if "FINAL ANSWER:" in raw: return raw.split("FINAL ANSWER:")[-1].strip() return raw.split("Final Answer:")[-1].strip() if "Final Answer:" in raw else raw class BasicAgent: def __init__(self): self.graph = build_graph() def __call__(self, question: str) -> str: messages = [HumanMessage(content=question)] output = self.graph.invoke({"messages": messages}) return parse_final_answer(output['messages'][-1].content) def run_agent_only(profile: gr.OAuthProfile | None): global cached_answers cached_answers = [] results_log = [] if not profile: return "Please login first.", None try: agent = BasicAgent() except Exception as e: return f"Agent Init Error: {e}", None try: questions = requests.get("https://agents-course-unit4-scoring.hf.space/questions", timeout=15).json() except Exception as e: return f"Error fetching questions: {e}", None with open("system_prompt.txt", "r") as f: system_prompt = f.read().strip() for item in questions: task_id = item.get("task_id") question = item.get("question") file_name = item.get("file_name") if not task_id or not question: continue try: user_message = question + (f"\n\nFile to use: {file_name}" if file_name else "") answer = agent(system_prompt + "\n\n" + user_message) cached_answers.append({"task_id": task_id, "submitted_answer": answer}) results_log.append({"Task ID": task_id, "Question": question, "Submitted Answer": answer}) except Exception as e: results_log.append({"Task ID": task_id, "Question": question, "Submitted Answer": f"AGENT ERROR: {e}"}) return "Agent finished. Click 'Submit Cached Answers' next.", pd.DataFrame(results_log) def submit_cached_answers(profile: gr.OAuthProfile | None): if not profile or not cached_answers: return "Nothing to submit. Run the agent first.", None payload = { "username": profile.username, "agent_code": f"https://huggingface.co/spaces/{os.getenv('SPACE_ID')}/tree/main", "answers": cached_answers } try: response = requests.post("https://agents-course-unit4-scoring.hf.space/submit", json=payload, timeout=60) result = response.json() score = result.get("score", "?") correct = result.get("correct_count", "?") total = result.get("total_attempted", "?") return f"Submission complete. Score: {score}% ({correct}/{total})", None except Exception as e: return f"Submission failed: {e}", None with gr.Blocks() as demo: gr.Markdown("""# Agent Evaluator 1. Login with Hugging Face 2. Run agent only 3. Submit answers""") gr.LoginButton() run_button = gr.Button("Run Agent") submit_button = gr.Button("Submit Cached Answers") status_box = gr.Textbox(label="Status", lines=4) table = gr.DataFrame(label="Answers Log") run_button.click(fn=run_agent_only, outputs=[status_box, table]) submit_button.click(fn=submit_cached_answers, outputs=[status_box, table])) if __name__ == "__main__": print("\n" + "-"*30 + " App Starting " + "-"*30) space_host_startup = os.getenv("SPACE_HOST") space_id_startup = os.getenv("SPACE_ID") if space_host_startup: print(f"✅ SPACE_HOST found: {space_host_startup}") print(f" Runtime URL: https://{space_host_startup}.hf.space") else: print("ℹ️ No SPACE_HOST found.") if space_id_startup: print(f"✅ SPACE_ID found: {space_id_startup}") print(f" Repo URL: https://huggingface.co/spaces/{space_id_startup}") else: print("ℹ️ No SPACE_ID found.") print("Launching Gradio Interface...") demo.launch(debug=True, share=False)