| exp_config = { | |
| 'main_config': { | |
| 'exp_name': 'TicTacToe-play-with-bot-AlphaZero', | |
| 'seed': 0, | |
| 'env': { | |
| 'env_id': 'TicTacToe-play-with-bot', | |
| 'board_size': 3, | |
| 'battle_mode': 'play_with_bot_mode', | |
| 'bot_action_type': 'v0', | |
| 'channel_last': False, | |
| 'collector_env_num': 8, | |
| 'evaluator_env_num': 5, | |
| 'n_evaluator_episode': 5, | |
| 'manager': { | |
| 'shared_memory': False | |
| }, | |
| 'agent_vs_human': False, | |
| 'prob_random_agent': 0, | |
| 'prob_expert_agent': 0, | |
| 'scale': True, | |
| 'alphazero_mcts_ctree': False, | |
| 'save_replay_gif': False, | |
| 'replay_path_gif': './replay_gif' | |
| }, | |
| 'policy': { | |
| 'on_policy': False, | |
| 'cuda': True, | |
| 'multi_gpu': False, | |
| 'bp_update_sync': True, | |
| 'traj_len_inf': False, | |
| 'model': { | |
| 'observation_shape': [3, 3, 3], | |
| 'action_space_size': 9, | |
| 'num_res_blocks': 1, | |
| 'num_channels': 16, | |
| 'fc_value_layers': [8], | |
| 'fc_policy_layers': [8] | |
| }, | |
| 'torch_compile': False, | |
| 'tensor_float_32': False, | |
| 'sampled_algo': False, | |
| 'gumbel_algo': False, | |
| 'update_per_collect': 50, | |
| 'model_update_ratio': 0.1, | |
| 'batch_size': 256, | |
| 'optim_type': 'Adam', | |
| 'learning_rate': 0.003, | |
| 'weight_decay': 0.0001, | |
| 'momentum': 0.9, | |
| 'grad_clip_value': 0.5, | |
| 'value_weight': 1.0, | |
| 'collector_env_num': 8, | |
| 'evaluator_env_num': 5, | |
| 'lr_piecewise_constant_decay': False, | |
| 'threshold_training_steps_for_final_lr': 500000, | |
| 'manual_temperature_decay': False, | |
| 'threshold_training_steps_for_final_temperature': 100000, | |
| 'fixed_temperature_value': 0.25, | |
| 'mcts': { | |
| 'num_simulations': 25 | |
| }, | |
| 'other': { | |
| 'replay_buffer': { | |
| 'replay_buffer_size': 1000000, | |
| 'save_episode': False | |
| } | |
| }, | |
| 'cfg_type': 'AlphaZeroPolicyDict', | |
| 'mcts_ctree': False, | |
| 'simulation_env_name': 'tictactoe', | |
| 'simulation_env_config_type': 'play_with_bot', | |
| 'board_size': 3, | |
| 'entropy_weight': 0.0, | |
| 'n_episode': 8, | |
| 'eval_freq': 2000 | |
| }, | |
| 'wandb_logger': { | |
| 'gradient_logger': False, | |
| 'video_logger': False, | |
| 'plot_logger': False, | |
| 'action_logger': False, | |
| 'return_logger': False | |
| } | |
| }, | |
| 'create_config': { | |
| 'env': { | |
| 'type': 'tictactoe', | |
| 'import_names': ['zoo.board_games.tictactoe.envs.tictactoe_env'] | |
| }, | |
| 'env_manager': { | |
| 'type': 'subprocess' | |
| }, | |
| 'policy': { | |
| 'type': 'alphazero', | |
| 'import_names': ['lzero.policy.alphazero'] | |
| }, | |
| 'collector': { | |
| 'type': 'episode_alphazero', | |
| 'import_names': ['lzero.worker.alphazero_collector'] | |
| }, | |
| 'evaluator': { | |
| 'type': 'alphazero', | |
| 'import_names': ['lzero.worker.alphazero_evaluator'] | |
| } | |
| } | |
| } | |