Spaces:
Running
Running
| #!/usr/bin/env python3 | |
| import os | |
| import sys | |
| import argparse | |
| from pathlib import Path | |
| from dotenv import load_dotenv | |
| from huggingface_hub import HfApi | |
| def main(): | |
| parser = argparse.ArgumentParser() | |
| parser.add_argument("--yes", action="store_true", help="Auto confirm deletion") | |
| args = parser.parse_args() | |
| REPO_ROOT = Path(__file__).resolve().parents[1] | |
| load_dotenv(REPO_ROOT / ".env") | |
| token = os.getenv("HF_TOKEN") | |
| repo_id = os.getenv("HF_SPACE_REPO", "ThinkcatLab/LiveHouse-TS") | |
| if not token: | |
| print("HF_TOKEN not found in .env", file=sys.stderr) | |
| return 1 | |
| print(f"Connecting to Hugging Face Space: {repo_id}") | |
| api = HfApi(token=token) | |
| # List all files in the repository | |
| try: | |
| files = api.list_repo_files(repo_id=repo_id, repo_type="space") | |
| except Exception as e: | |
| print(f"Failed to list files: {e}", file=sys.stderr) | |
| return 1 | |
| allowed_dirs = { | |
| "aggregates", | |
| "arima", | |
| "chronos_bolt", | |
| "chronos_2", | |
| "ets", | |
| "moirai_2_0", | |
| "moving_average", | |
| "seasonal_naive", | |
| "sundial", | |
| "tabpfn_ts", | |
| "timesfm_2_5", | |
| "tirex", | |
| "toto_1_0", | |
| } | |
| allowed_files = { | |
| "online_status.json", | |
| "dataset_properties.csv", | |
| "dataset_eval_intervals.json", | |
| "eval_history.jsonl", | |
| "eval_state.json", | |
| "baseline_rank_history.csv", | |
| "baseline_rank_history_daily.csv", | |
| "baseline_rank_history_weekly.csv", | |
| "model_allowlist.json", | |
| } | |
| files_to_delete = [] | |
| for f in files: | |
| if not f.startswith("results/"): | |
| continue | |
| parts = f.split("/") | |
| if len(parts) < 2: | |
| continue | |
| # check if it is a file directly under results/ | |
| if len(parts) == 2: | |
| filename = parts[1] | |
| if filename in {"evaluation_metrics.jsonl", "evaluation_metrics_canonical.jsonl"}: | |
| files_to_delete.append(f) | |
| elif filename not in allowed_files: | |
| files_to_delete.append(f) | |
| else: | |
| # it is inside a subdirectory under results/ | |
| sub_dir = parts[1] | |
| if sub_dir == "aggregates" and ( | |
| parts[-1].endswith(".zip") | |
| or parts[-1] == "evaluation_metrics_canonical.jsonl" | |
| or "canonical_all_history" in parts | |
| ): | |
| files_to_delete.append(f) | |
| elif sub_dir == "saved_evals" or "forecasts" in parts: | |
| files_to_delete.append(f) | |
| elif sub_dir not in allowed_dirs: | |
| files_to_delete.append(f) | |
| if not files_to_delete: | |
| print("No extra files to delete on remote Space.") | |
| return 0 | |
| print(f"Found {len(files_to_delete)} files to delete from Hugging Face Space:") | |
| for f in files_to_delete[:20]: | |
| print(f" - {f}") | |
| if len(files_to_delete) > 20: | |
| print(f" ... and {len(files_to_delete) - 20} more.") | |
| if not args.yes: | |
| confirm = input("Proceed with deletion? (y/n): ").strip().lower() | |
| if confirm != "y": | |
| print("Cancelled.") | |
| return 0 | |
| # Delete the files using api.delete_file or commit operations | |
| from huggingface_hub import CommitOperationDelete | |
| operations = [CommitOperationDelete(path_in_repo=f) for f in files_to_delete] | |
| # We batch commit in chunks of 500 files to avoid payload limits | |
| chunk_size = 500 | |
| for i in range(0, len(operations), chunk_size): | |
| chunk = operations[i:i+chunk_size] | |
| print(f"Deleting chunk {i // chunk_size + 1} ({len(chunk)} files)...") | |
| try: | |
| api.create_commit( | |
| repo_id=repo_id, | |
| repo_type="space", | |
| operations=chunk, | |
| commit_message=f"Clean remote results: delete extra model files (part {i // chunk_size + 1})", | |
| ) | |
| except Exception as e: | |
| print(f"Failed to delete chunk: {e}", file=sys.stderr) | |
| return 1 | |
| print("Deletion completed successfully.") | |
| return 0 | |
| if __name__ == "__main__": | |
| sys.exit(main()) | |