This project provides a set of translators to convert OpenAI Gym environments into text-based environments. It is designed to investigate the capabilities of large language models in decision-making tasks within these text-based environments.
The Ant env is ready. Test with python main_reflexion.py --env_name Ant-v4 --init_summarizer ant_init_translator --curr_summarizer ant_basic_translator --decider naive_actor --prompt_level 3 --num_trails 1 --distiller guide_generator
The Ant env is ready. Test with
python main_reflexion.py --env_name Ant-v4 --init_summarizer ant_init_translator --curr_summarizer ant_basic_translator --decider naive_actor --prompt_level 3 --num_trails 1 --distiller guide_generator