Spaces:
Sleeping
Sleeping
| import random | |
| import argparse | |
| from src.tasks import EasyTask, MediumTask, HardTask | |
| from src.agent import DeterministicAgent | |
| def run_evaluation(base_seed=None, silent=False): | |
| if base_seed is None: | |
| base_seed = random.randint(1000, 99999) | |
| random.seed(base_seed) | |
| if not silent: | |
| print("==================================================") | |
| print(f"=== Smart Traffic Eval (Seed: {base_seed}) ===") | |
| agent = DeterministicAgent() | |
| tasks = { | |
| "Easy": EasyTask(), | |
| "Medium": MediumTask(), | |
| "Hard": HardTask() | |
| } | |
| results = {} | |
| total_score = 0.0 | |
| for level, task in tasks.items(): | |
| task_seed = base_seed + list(tasks.keys()).index(level) * 999 | |
| state = task.reset(seed=task_seed) | |
| done = False | |
| steps = 0 | |
| total_reward = 0.0 | |
| while not done: | |
| action_idx = agent.get_action(state) | |
| result = task.step(action_idx) | |
| state = result.state | |
| reward = result.reward | |
| done = result.done | |
| total_reward += reward | |
| steps += 1 | |
| if steps > 500: | |
| break | |
| score = task.evaluate() | |
| total_score += score | |
| results[level] = score | |
| info = result.info | |
| if not silent: | |
| print(f"[{level}] Steps: {steps} | Total Reward: {total_reward:.2f}") | |
| print(f" Cleared: {info['total_cleared']} | Avg Wait/Car: {info['avg_waiting_time']:.1f} | Emg Handled: {info['emergencies_handled']}") | |
| print(f" Final Level Score (0-1): {score:.3f}") | |
| avg_score = total_score / len(tasks) | |
| results["Overall"] = avg_score | |
| if not silent: | |
| print(f"==================================================") | |
| print(f"Overall Average Score: {avg_score:.3f} / 1.000") | |
| print(f"==================================================\n") | |
| return results | |
| if __name__ == "__main__": | |
| parser = argparse.ArgumentParser(description="Evaluate the Smart Traffic Agent") | |
| parser.add_argument("--seed", type=int, default=None, help="Fix the RNG seed for reproducible testing") | |
| args = parser.parse_args() | |
| run_evaluation(base_seed=args.seed) | |