Spaces:
Sleeping
Sleeping
| import json | |
| import os | |
| import tempfile | |
| from pathlib import Path | |
| import gradio as gr | |
| import pandas as pd | |
| import requests | |
| from agent import GaiaAgent | |
| from answer_normalize import normalize_answer | |
| DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space" | |
| CACHE_FILENAME = "gaia_answers_cache.json" | |
| def _cache_path() -> Path: | |
| return Path(__file__).resolve().parent / CACHE_FILENAME | |
| def _question_cache_tag(question: str) -> str: | |
| """Bind cached answers to question text so task_id alone cannot serve stale rows.""" | |
| s = " ".join(str(question).split()) | |
| return s[:280] | |
| def _load_cache() -> dict[str, dict]: | |
| p = _cache_path() | |
| if not p.is_file(): | |
| return {} | |
| try: | |
| raw = json.loads(p.read_text(encoding="utf-8")) | |
| except json.JSONDecodeError: | |
| return {} | |
| if not isinstance(raw, dict): | |
| return {} | |
| out: dict[str, dict] = {} | |
| for k, v in raw.items(): | |
| if not isinstance(k, str): | |
| continue | |
| if isinstance(v, dict) and isinstance(v.get("a"), str) and isinstance(v.get("qtag"), str): | |
| out[k] = v | |
| # Legacy format task_id -> plain string (unsafe if questions rotate): ignore. | |
| return out | |
| def _save_cache(cache: dict[str, dict]) -> None: | |
| _cache_path().write_text(json.dumps(cache, indent=2), encoding="utf-8") | |
| def _cache_get(cache: dict[str, dict], task_id: str, question_text: str) -> str | None: | |
| entry = cache.get(str(task_id)) | |
| if not entry: | |
| return None | |
| if entry.get("qtag") != _question_cache_tag(question_text): | |
| return None | |
| return entry.get("a") | |
| def _cache_set( | |
| cache: dict[str, dict], task_id: str, question_text: str, answer: str | |
| ) -> None: | |
| cache[str(task_id)] = { | |
| "qtag": _question_cache_tag(question_text), | |
| "a": answer, | |
| } | |
| def _download_attachment(api_url: str, task_id: str, file_name: str) -> str | None: | |
| """Save task attachment to a temp file; return path or None.""" | |
| if not file_name or not str(file_name).strip(): | |
| return None | |
| url = f"{api_url}/files/{task_id}" | |
| try: | |
| r = requests.get( | |
| url, | |
| timeout=120, | |
| allow_redirects=True, | |
| headers={ | |
| "User-Agent": "GAIA-Agent/1.0 (HuggingFace-Space; +https://huggingface.co)" | |
| }, | |
| ) | |
| except requests.RequestException: | |
| return None | |
| if r.status_code != 200: | |
| return None | |
| ctype = (r.headers.get("Content-Type") or "").lower() | |
| if "application/json" in ctype: | |
| try: | |
| data = r.json() | |
| if isinstance(data, dict) and data.get("detail"): | |
| return None | |
| except json.JSONDecodeError: | |
| pass | |
| suffix = Path(file_name).suffix or "" | |
| fd, path = tempfile.mkstemp(suffix=suffix, prefix=f"gaia_{task_id[:8]}_") | |
| try: | |
| with os.fdopen(fd, "wb") as f: | |
| f.write(r.content) | |
| except OSError: | |
| return None | |
| return path | |
| def run_and_submit_all(profile: gr.OAuthProfile | None): | |
| space_id = os.getenv("SPACE_ID") | |
| use_cache = os.getenv("GAIA_USE_CACHE", "0").lower() in ("1", "true", "yes") | |
| 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 = os.getenv("GAIA_API_URL", DEFAULT_API_URL) | |
| questions_url = f"{api_url}/questions" | |
| submit_url = f"{api_url}/submit" | |
| try: | |
| agent = GaiaAgent() | |
| 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(agent_code) | |
| print(f"Fetching questions from: {questions_url}") | |
| try: | |
| response = requests.get(questions_url, timeout=60) | |
| 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 requests.exceptions.RequestException as e: | |
| return f"Error fetching questions: {e}", None | |
| except json.JSONDecodeError as e: | |
| return f"Error decoding server response for questions: {e}", None | |
| cache = _load_cache() if use_cache else {} | |
| 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") | |
| file_name = item.get("file_name") or "" | |
| if not task_id or question_text is None: | |
| print(f"Skipping item with missing task_id or question: {item}") | |
| continue | |
| cache_key = str(task_id) | |
| cached_raw = _cache_get(cache, cache_key, str(question_text)) if use_cache else None | |
| if cached_raw is not None: | |
| submitted_answer = normalize_answer( | |
| cached_raw, context_question=str(question_text) | |
| ) | |
| print(f"Cache hit for {task_id}") | |
| else: | |
| local_path: str | None = None | |
| try: | |
| if file_name and str(file_name).strip(): | |
| local_path = _download_attachment(api_url, str(task_id), str(file_name)) | |
| if local_path: | |
| print(f"Downloaded attachment for {task_id} -> {local_path}") | |
| submitted_answer = agent( | |
| str(question_text), | |
| attachment_path=local_path, | |
| task_id=str(task_id), | |
| ) | |
| submitted_answer = normalize_answer( | |
| submitted_answer, context_question=str(question_text) | |
| ) | |
| if use_cache: | |
| _cache_set( | |
| cache, | |
| cache_key, | |
| str(question_text), | |
| submitted_answer | |
| if isinstance(submitted_answer, str) | |
| else str(submitted_answer), | |
| ) | |
| _save_cache(cache) | |
| except Exception as e: | |
| print(f"Error running agent on task {task_id}: {e}") | |
| submitted_answer = f"AGENT ERROR: {e}" | |
| finally: | |
| if local_path and Path(local_path).is_file(): | |
| try: | |
| Path(local_path).unlink(missing_ok=True) | |
| except OSError: | |
| pass | |
| 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, | |
| } | |
| ) | |
| 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, | |
| } | |
| status_update = ( | |
| f"Agent finished. Submitting {len(answers_payload)} answers for user '{username}'..." | |
| ) | |
| print(status_update) | |
| print(f"Submitting {len(answers_payload)} answers to: {submit_url}") | |
| try: | |
| response = requests.post(submit_url, json=submission_data, timeout=600) | |
| 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}." | |
| try: | |
| error_json = e.response.json() | |
| error_detail += f" Detail: {error_json.get('detail', e.response.text)}" | |
| except json.JSONDecodeError: | |
| error_detail += f" Response: {e.response.text[:500]}" | |
| status_message = f"Submission Failed: {error_detail}" | |
| print(status_message) | |
| return status_message, pd.DataFrame(results_log) | |
| except requests.exceptions.Timeout: | |
| status_message = "Submission Failed: The request timed out." | |
| print(status_message) | |
| return status_message, pd.DataFrame(results_log) | |
| except requests.exceptions.RequestException as e: | |
| status_message = f"Submission Failed: Network error - {e}" | |
| print(status_message) | |
| return status_message, pd.DataFrame(results_log) | |
| except Exception as e: | |
| status_message = f"An unexpected error occurred during submission: {e}" | |
| print(status_message) | |
| return status_message, pd.DataFrame(results_log) | |
| def crypto_btc_price() -> str: | |
| """Optional demo: live BTC/USD (not used for GAIA scoring).""" | |
| try: | |
| r = requests.get( | |
| "https://api.coingecko.com/api/v3/simple/price", | |
| params={"ids": "bitcoin", "vs_currencies": "usd"}, | |
| timeout=20, | |
| ) | |
| r.raise_for_status() | |
| data = r.json() | |
| usd = data.get("bitcoin", {}).get("usd") | |
| return f"Bitcoin (BTC) ~ ${usd:,.2f} USD (CoinGecko public API)." | |
| except Exception as e: | |
| return f"Could not fetch price: {e}" | |
| with gr.Blocks() as demo: | |
| gr.Markdown("# GAIA Unit 4 — Agent Evaluation Runner") | |
| gr.Markdown( | |
| """ | |
| **Instructions** | |
| 1. Duplicate this Space from the course template (or push this repo) and set **Secrets**: `HF_TOKEN` (read access to Inference). | |
| 2. Optional env vars: `GAIA_TEXT_MODEL`, `GAIA_ASR_MODEL`, `GAIA_VISION_MODEL`, `GAIA_API_URL`, `GAIA_USE_CACHE` (default **`0`** — answers are keyed by `task_id` **and** question text; set `1` only to speed re-runs). | |
| 3. Log in with Hugging Face below (username is used for the leaderboard). | |
| 4. Run **Evaluate & Submit** to answer all questions and post scores. | |
| Attachment tasks download `GET /files/{task_id}` automatically when `file_name` is set. | |
| --- | |
| **Crypto demo (optional):** unrelated to GAIA; quick BTC spot check. | |
| """ | |
| ) | |
| gr.LoginButton() | |
| with gr.Tab("GAIA evaluation"): | |
| run_button = gr.Button("Run Evaluation & Submit All Answers") | |
| status_output = gr.Textbox( | |
| label="Run Status / Submission Result", lines=6, 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], | |
| ) | |
| with gr.Tab("Crypto intelligence (demo)"): | |
| gr.Markdown( | |
| "This tab does not affect GAIA scores. It demonstrates a simple public market data fetch." | |
| ) | |
| cp_btn = gr.Button("Fetch BTC / USD") | |
| cp_out = gr.Textbox(label="Output", interactive=False) | |
| cp_btn.click(fn=crypto_btc_price, outputs=cp_out) | |
| 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}") | |
| else: | |
| print("SPACE_HOST not set (local run?).") | |
| if space_id_startup: | |
| print(f"SPACE_ID found: {space_id_startup}") | |
| print(f"Repo tree URL: https://huggingface.co/spaces/{space_id_startup}/tree/main") | |
| else: | |
| print("SPACE_ID not set (local run?).") | |
| print("-" * 62 + "\n") | |
| demo.launch(debug=True, share=False) | |