feat: add run_multienv_o3.py script for multi-environment evaluation

- Introduced a new script `run_multienv_o3.py` to facilitate end-to-end evaluation across multiple environments.
- Implemented command-line argument parsing for various configurations including environment settings, logging levels, and AWS parameters.
- Integrated signal handling for graceful shutdown of environments and processes.
- Enhanced logging capabilities for better traceability during execution.
- Maintained existing logic from previous scripts while introducing new functionalities for improved evaluation processes.

run_multienv_uitars15.py

@@ -207,7 +207,6 @@ def run_env_tasks(task_queue: Queue, args: argparse.Namespace, shared_scores: li
             max_tokens=args.max_tokens,
             top_p=args.top_p,
             temperature=args.temperature,
-            
             max_trajectory_length=args.max_trajectory_length,
             max_image_history_length=args.max_image_history_length,
             use_thinking=args.use_thinking,