import os import gradio as gr import requests import inspect import pandas as pd import wikipedia import arxiv from duckduckgo_search import DDGS import json from langgraph.graph import StateGraph from typing import TypedDict, Optional import openai # (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 ------ # 定義狀態 class QueryState(TypedDict): question: str query_type: str wiki_result: Optional[str] web_result: Optional[str] arxiv_result: Optional[str] final_answer: Optional[str] # 節點定義 def analyze_question(state: QueryState) -> QueryState: new_state = state.copy() q = new_state["question"].lower() if any(x in q for x in ["arxiv", "paper", "journal", "study", "research", "doi"]): new_state['query_type'] = "academic" elif any(x in q for x in ["today", "latest", "news", "current", "recent", "2023", "2024"]): new_state['query_type'] = "current" elif any(x in q for x in ["album", "who", "how many", "what", "which", "name", "country", "city", "code", "recipient", "team", "athlete", "author", "number"]): new_state['query_type'] = "fact" else: new_state['query_type'] = "other" return new_state def extract_query_with_gpt(question: str) -> str: prompt = '''You are a search query generator.\nGiven a user question, extract the most relevant and concise English search query or keywords that would help find the answer in Wikipedia or Google.\nOnly output the search query, do not explain or translate.\n\nExamples:\nQuestion: Who won the Nobel Prize in Physics in 2020?\nSearch query: Nobel Prize Physics 2020 winner\n\nQuestion: How many studio albums were published by Mercedes Sosa between 2000 and 2009?\nSearch query: Mercedes Sosa studio albums 2000-2009\n\nQuestion: What is the capital city of Mongolia?\nSearch query: Mongolia capital city\n\nQuestion: {question}\nSearch query:'''.format(question=question) try: response = openai.ChatCompletion.create( model="gpt-4o", messages=[{"role": "user", "content": prompt}], max_tokens=32, temperature=0.2, ) return response.choices[0].message.content.strip() except Exception as e: return question # fallback: 用原題目 def search_wikipedia(state: QueryState) -> QueryState: new_state = state.copy() if new_state['query_type'] != 'fact': new_state['wiki_result'] = "Not supported" return new_state query = extract_query_with_gpt(new_state['question']) try: new_state['wiki_result'] = wikipedia.summary(query, sentences=2) except Exception as e: new_state['wiki_result'] = f"[Wikipedia無結果] {e}" return new_state def search_web(state: QueryState) -> QueryState: new_state = state.copy() if new_state['query_type'] != 'fact': new_state['web_result'] = "Not supported" return new_state query = extract_query_with_gpt(new_state['question']) try: with DDGS() as ddgs: results = list(ddgs.text(query, max_results=2)) if results: new_state['web_result'] = results[0]['body'] else: new_state['web_result'] = "[DuckDuckGo無結果]" except Exception as e: new_state['web_result'] = f"[DuckDuckGo錯誤] {e}" return new_state def search_arxiv(state: QueryState) -> QueryState: new_state = state.copy() if new_state['query_type'] != 'fact': new_state['arxiv_result'] = "Not supported" return new_state query = extract_query_with_gpt(new_state['question']) try: client = arxiv.Client() search = arxiv.Search(query=query, max_results=1, sort_by=arxiv.SortCriterion.Relevance) results = list(client.results(search)) if results: paper = results[0] new_state['arxiv_result'] = f"arXiv: {paper.title}\n{paper.summary[:200]}..." else: new_state['arxiv_result'] = "[arXiv無結果]" except Exception as e: new_state['arxiv_result'] = f"[arXiv錯誤] {e}" return new_state def synthesize_answer(state: QueryState) -> QueryState: new_state = state.copy() if new_state['query_type'] != 'fact': new_state['final_answer'] = "Not supported" return new_state prompt = f'''You are an expert answer extractor.\nGiven the following search results from Wikipedia, DuckDuckGo, and arXiv, extract and output ONLY the final, most accurate answer to the original question.\nDo not explain, do not repeat the question, do not add any extra text.\nIf the answer is a number, only output the number. If it is a name, only output the name. If the answer is a list, only output the list in the required format. If you cannot find the answer, output None.\n\nExamples:\nOriginal question: How many studio albums were published by Mercedes Sosa between 2000 and 2009?\n[Wikipedia]\nMercedes Sosa released three studio albums between 2000 and 2009.\n[DuckDuckGo]\nMercedes Sosa's discography includes albums released in 2000, 2002, and 2005.\n[arXiv]\nNo relevant results.\nFinal answer: 3\n\nOriginal question: What is the capital city of Mongolia?\n[Wikipedia]\nThe capital city of Mongolia is Ulaanbaatar.\n[DuckDuckGo]\nUlaanbaatar is the capital of Mongolia.\n[arXiv]\nNo relevant results.\nFinal answer: Ulaanbaatar\n\nOriginal question: {new_state['question']}\n[Wikipedia]\n{new_state['wiki_result']}\n[DuckDuckGo]\n{new_state['web_result']}\n[arXiv]\n{new_state['arxiv_result']}\nFinal answer:''' try: response = openai.ChatCompletion.create( model="gpt-4o", messages=[{"role": "user", "content": prompt}], max_tokens=64, temperature=0.2, ) new_state['final_answer'] = response.choices[0].message.content.strip() except Exception as e: new_state['final_answer'] = "[LLM後處理錯誤]" return new_state # 條件分支 def route_from_analyze(state: QueryState): if state['query_type'] == "fact": return "wiki" elif state['query_type'] == "academic": return "arxiv" elif state['query_type'] == "current": return "web" else: return "synthesize" # 構建 LangGraph class QueryAgent: def __init__(self): print("QueryAgent (LangGraph) initialized.") self.graph = StateGraph(QueryState) self.graph.add_node("analyze", analyze_question) self.graph.add_node("wiki", search_wikipedia) self.graph.add_node("web", search_web) self.graph.add_node("arxiv", search_arxiv) self.graph.add_node("synthesize", synthesize_answer) self.graph.add_conditional_edges( "analyze", route_from_analyze, { "wiki": "wiki", "web": "web", "arxiv": "arxiv", "synthesize": "synthesize" } ) self.graph.add_edge("wiki", "web") self.graph.add_edge("web", "arxiv") self.graph.add_edge("arxiv", "synthesize") self.graph.set_entry_point("analyze") self.compiled_graph = self.graph.compile() def __call__(self, question: str) -> str: state: QueryState = { "question": question, "query_type": "", "wiki_result": None, "web_result": None, "arxiv_result": None, "final_answer": None } result_state = self.compiled_graph.invoke(state) return result_state["final_answer"] or "無法產生答案" def run_and_submit_all( profile: gr.OAuthProfile | None): """ Fetches all questions, runs the QueryAgent 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 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 = QueryAgent() 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 json.decoder.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 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}) 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) results_df.to_csv("my_agent_answers.csv", index=False) 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 json.decoder.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) results_df.to_csv("my_agent_answers.csv", index=False) 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) results_df.to_csv("my_agent_answers.csv", index=False) 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) results_df.to_csv("my_agent_answers.csv", index=False) 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) results_df.to_csv("my_agent_answers.csv", index=False) return status_message, results_df # --- Build Gradio Interface using Blocks --- with gr.Blocks() as demo: gr.Markdown("# Query 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] ) 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 Query Agent Evaluation...") demo.launch(debug=True, share=False)