#!/bin/env python """Little command-line utility for probing a model on disk.""" import argparse from pathlib import Path from invokeai.backend.model_manager import InvalidModelConfigException, ModelProbe parser = argparse.ArgumentParser(description="Probe model type") parser.add_argument( "model_path", type=Path, nargs="+", ) args = parser.parse_args() for path in args.model_path: try: info = ModelProbe.probe(path) print(f"{path}:{info.model_dump_json(indent=4)}") except InvalidModelConfigException as exc: print(exc)