import matplotlib matplotlib.use('Agg') import os import shutil import tarfile import tempfile from datetime import datetime, timedelta, timezone from email.utils import parseaddr from pathlib import Path import gradio as gr import requests from agenteval import ( process_eval_logs, upload_folder_to_hf, upload_summary_to_hf, ) from agenteval.models import EvalResult from agenteval.leaderboard.upload import sanitize_path_component from datasets import Dataset, DatasetDict, VerificationMode, load_dataset from datasets.data_files import EmptyDatasetError from huggingface_hub import HfApi from content import ( CITATION_BUTTON_LABEL, CITATION_BUTTON_TEXT, format_error, format_log, format_warning, ) # --- Constants and Configuration --- LOCAL_DEBUG = not (os.environ.get("system") == "spaces") CONFIG_NAME = "1.0.0-dev1" # This corresponds to 'config' in LeaderboardViewer IS_INTERNAL = os.environ.get("IS_INTERNAL", "false").lower() == "true" OWNER = "allenai" PROJECT_NAME = "asta-bench" + ("-internal" if IS_INTERNAL else "") SUBMISSION_DATASET = f"{OWNER}/{PROJECT_NAME}-submissions" SUBMISSION_DATASET_PUBLIC = f"{OWNER}/{PROJECT_NAME}-submissions-public" CONTACT_DATASET = f"{OWNER}/{PROJECT_NAME}-contact-info" RESULTS_DATASET = f"{OWNER}/{PROJECT_NAME}-results" # This is the repo_id for LeaderboardViewer LEADERBOARD_PATH = f"{OWNER}/{PROJECT_NAME}-leaderboard" if LOCAL_DEBUG: DATA_DIR = os.path.join(os.path.dirname(__file__), "data", CONFIG_NAME) else: DATA_DIR = "/home/user/data/" + CONFIG_NAME EXTRACTED_DATA_DIR = os.path.join(DATA_DIR, "extracted") api = HfApi() MAX_UPLOAD_BYTES = 100 * 1024**2 AGENTEVAL_MANIFEST_NAME = "agenteval.json" os.makedirs(EXTRACTED_DATA_DIR, exist_ok=True) # --- Global State for Viewers (simple caching) --- CACHED_VIEWERS = {} CACHED_TAG_MAPS = {} # --- Submission Logic (largely unchanged from original, ensure EvalResult and other deps are fine) --- def try_load_dataset_submission(*args, **kwargs) -> DatasetDict: # Renamed to avoid conflict if LV has one try: return load_dataset(*args, **kwargs) except EmptyDatasetError: return DatasetDict() except ValueError: # Handles cases where dataset is empty or ill-formed return DatasetDict() def checked_upload_folder( api_hf: HfApi, # Renamed to avoid conflict with global api folder_path: str, repo_id: str, config_name_ul: str, # Renamed split_ul: str, # Renamed submission_name_ul: str, # Renamed ) -> str: total = 0 for root, _, files in os.walk(folder_path): for f_ul in files: # Renamed total += os.path.getsize(os.path.join(root, f_ul)) if total > MAX_UPLOAD_BYTES: raise ValueError( f"Upload too large: exceeds {MAX_UPLOAD_BYTES // (1024**2)} MB limit." ) return upload_folder_to_hf( api=api_hf, # Use renamed parameter folder_path=folder_path, repo_id=repo_id, config_name=config_name_ul, split=split_ul, submission_name=submission_name_ul, ) def add_new_eval( val_or_test: str, agent_name: str | None, agent_description: str, agent_url: str, openness: str | None, degree_of_control: str | None, path_to_file: tempfile._TemporaryFileWrapper | None, username: str, mail: str, profile: gr.OAuthProfile, # We need global eval_results for checks; this might need rethinking if it's purely display driven now # For now, let's assume we still load it for submission checks ): # Load current eval_results for submission checks # This is a bit redundant if display part reloads it, but submission needs its own consistent view current_eval_results_for_submission = try_load_dataset_submission( RESULTS_DATASET, CONFIG_NAME, download_mode="force_redownload", # Or a less aggressive mode verification_mode=VerificationMode.NO_CHECKS, trust_remote_code=True, ) if not agent_name: return format_warning("Please provide an agent name.") submission_time = datetime.now(timezone.utc) if not username or username.strip() == "": username = profile.username # Default to HF username # User account age check try: user_data_resp = requests.get(f"https://huggingface.co/api/users/{profile.username}/overview") user_data_resp.raise_for_status() creation_date_str = user_data_resp.json()["createdAt"] created_at = datetime.strptime(creation_date_str, "%Y-%m-%dT%H:%M:%S.%fZ").replace(tzinfo=timezone.utc) if submission_time - created_at < timedelta(days=60): return format_error("This account is not authorized to submit here (account too new).") except Exception as e: print(f"Error checking user account age: {e}") return format_error("Could not verify account age. Please try again later.") # Submission frequency check contact_infos = try_load_dataset_submission( CONTACT_DATASET, CONFIG_NAME, download_mode="force_redownload", verification_mode=VerificationMode.NO_CHECKS, trust_remote_code=True ) user_submission_dates = sorted( datetime.fromisoformat(row["submit_time"]) for row in contact_infos.get(val_or_test, []) if row["username_auth"] == profile.username ) if user_submission_dates and (submission_time - user_submission_dates[-1] < timedelta(days=1)): return format_error("You already submitted once in the last 24h for this split; please try again later.") # Email validation _, parsed_mail = parseaddr(mail) if "@" not in parsed_mail: return format_warning("Please provide a valid email address.") # Duplicate submission check if val_or_test in current_eval_results_for_submission and len(current_eval_results_for_submission[val_or_test]) > 0: existing_submissions = current_eval_results_for_submission[val_or_test].to_dict().get("submission", []) for sub_item in existing_submissions: if (sub_item.get("agent_name", "").lower() == agent_name.lower() and sub_item.get("username", "").lower() == username.lower()): return format_warning("This agent name by this user has already been submitted to this split.") if path_to_file is None: return format_warning("Please attach a .tar.gz file.") safe_username = sanitize_path_component(username) safe_agent_name = sanitize_path_component(agent_name) extracted_dir = os.path.join(EXTRACTED_DATA_DIR, f"{safe_username}_{safe_agent_name}") # File extraction if not LOCAL_DEBUG: try: if os.path.exists(extracted_dir): shutil.rmtree(extracted_dir) os.makedirs(extracted_dir, exist_ok=True) with tarfile.open(path_to_file.name, "r:gz") as tar: members_extracted = 0 for member in tar.getmembers(): if not member.isreg(): continue fname = os.path.basename(member.name) if not fname or fname.startswith("."): continue fobj = tar.extractfile(member) if not fobj: continue with open(os.path.join(extracted_dir, fname), "wb") as out: out.write(fobj.read()) members_extracted +=1 if members_extracted == 0: return format_error("Submission tarball is empty or contains no valid files.") except Exception as e: return format_error(f"Error extracting file: {e}. Ensure it's a valid .tar.gz.") else: print("mock extracted file", flush=True) submission_name = f"{safe_username}_{safe_agent_name}_{submission_time.strftime('%Y-%m-%d_%H-%M-%S')}" # 1. Upload raw (unscored) submission files if not LOCAL_DEBUG: try: checked_upload_folder(api, extracted_dir, SUBMISSION_DATASET, CONFIG_NAME, val_or_test, submission_name) except ValueError as e: return format_error(str(e)) except Exception as e: return format_error(f"Failed to upload raw submission: {e}") else: print("mock uploaded raw submission", flush=True) # 2. Save contact information contact_info = { "agent_name": agent_name, "agent_description": agent_description, "url": agent_url, "username": username, "username_auth": profile.username, "mail": mail, "submit_time": submission_time.isoformat(), } if val_or_test in contact_infos: contact_infos[val_or_test] = contact_infos[val_or_test].add_item(contact_info) else: contact_infos[val_or_test] = Dataset.from_list([contact_info]) if not LOCAL_DEBUG: try: contact_infos.push_to_hub(CONTACT_DATASET, config_name=CONFIG_NAME) except Exception as e: return format_warning(f"Submission recorded, but contact info failed to save: {e}") else: print("mock uploaded contact info", flush=True) # 3. Process and score the submission eval_result_obj = None # Define to avoid NameError try: json_path = Path(extracted_dir) / AGENTEVAL_MANIFEST_NAME if not json_path.exists(): return format_error(f"Missing manifest {AGENTEVAL_MANIFEST_NAME} in submission.") eval_result_obj = EvalResult.model_validate_json(json_path.read_text(encoding="utf-8")) if eval_result_obj.suite_config.version != CONFIG_NAME: return format_error(f"Suite version mismatch: expected {CONFIG_NAME}, got {eval_result_obj.suite_config.version}.") if eval_result_obj.split != val_or_test: return format_error(f"Split mismatch: expected {val_or_test}, got {eval_result_obj.split}.") # Re-compute results from logs for integrity eval_result_obj.results = process_eval_logs(extracted_dir)[0] # Assuming process_eval_logs returns a tuple/list eval_result_obj.save_json(str(json_path)) # Save the re-processed manifest except Exception as e: return format_error(f"Error scoring submission: {e}. Check manifest and log files.") # 4. Upload scored submission files logs_url_private_val, logs_url_public_val = None, None scored_submission_name = f"{submission_name}_scored" if not LOCAL_DEBUG: try: logs_url_private_val = checked_upload_folder(api, extracted_dir, SUBMISSION_DATASET, CONFIG_NAME, val_or_test, scored_submission_name) if val_or_test == "validation" and not IS_INTERNAL: # Public copy for validation logs_url_public_val = checked_upload_folder(api, extracted_dir, SUBMISSION_DATASET_PUBLIC, CONFIG_NAME, val_or_test, scored_submission_name) except ValueError as e: return format_error(str(e)) except Exception as e: return format_error(f"Failed to upload scored submission: {e}") else: print("mock uploaded scored submission", flush=True) # Update EvalResult with submission details eval_result_obj.submission.agent_name = agent_name eval_result_obj.submission.agent_description = agent_description eval_result_obj.submission.agent_url = agent_url eval_result_obj.submission.openness = openness eval_result_obj.submission.degree_of_control = degree_of_control eval_result_obj.submission.username = username eval_result_obj.submission.submit_time = submission_time eval_result_obj.submission.logs_url = logs_url_private_val eval_result_obj.submission.logs_url_public = logs_url_public_val # 5. Upload summary statistics to RESULTS_DATASET (for the leaderboard) if not LOCAL_DEBUG: try: upload_summary_to_hf(api, eval_result_obj, RESULTS_DATASET, CONFIG_NAME, val_or_test, scored_submission_name) except Exception as e: return format_error(f"Failed to upload summary results to leaderboard: {e}") else: print("mock uploaded results to lb", flush=True) # Invalidate viewer cache for the split that was updated if val_or_test in CACHED_VIEWERS: del CACHED_VIEWERS[val_or_test] if val_or_test in CACHED_TAG_MAPS: del CACHED_TAG_MAPS[val_or_test] return format_log( f"Agent '{agent_name}' submitted successfully by '{username}' to '{val_or_test}' split. " "Please refresh the leaderboard in a few moments. It may take some time for changes to propagate." ) # --- Submission Accordion --- with gr.Blocks() as demo: gr.Markdown(f"## 🚀 Submit a new agent for evaluation", elem_id="markdown-text") with gr.Row(): with gr.Column(): level_of_test_radio = gr.Radio(["validation", "test"], value="validation", label="Split") agent_name_tb = gr.Textbox(label="Agent Name") agent_desc_tb = gr.Textbox(label="Agent Description") agent_url_tb = gr.Textbox(label="URL to Agent Information") openness_radio = gr.Radio(["Open Source","Open Source Open Weights", "API Available", "Closed"], value=None, label="Openness of Agent") degree_of_control_radio = gr.Radio(["Standard","Custom with Standard Search", "Fully Custom"], value=None, label="Agent Tooling") with gr.Column(): username_tb = gr.Textbox(label="Organization or User Name (Defaults to HF username)") mail_tb = gr.Textbox(label="Contact Email (Private, for submission issues)") file_upload_comp = gr.File( label="Submission File (.tar.gz ...)", # Shortened for brevity file_types=[".gz", ".tar.gz"] ) with gr.Row(): gr.LoginButton() submit_eval_button = gr.Button("Submit Evaluation") submission_result = gr.Markdown() submit_eval_button.click( add_new_eval, [ level_of_test_radio, agent_name_tb, agent_desc_tb, agent_url_tb, openness_radio, degree_of_control_radio, file_upload_comp, username_tb, mail_tb ], submission_result, ) with gr.Accordion("📙 Citation", open=False): gr.Textbox(value=CITATION_BUTTON_TEXT, label=CITATION_BUTTON_LABEL, elem_id="citation-button-main", interactive=False)