gomoku_rl package

Subpackages

• gomoku_rl.policy package
  • Submodules
    • gomoku_rl.policy.base module
      • Policy
    • gomoku_rl.policy.common module
      • get_optimizer()
      • make_critic()
      • make_dataset_naive()
      • make_dqn_actor()
      • make_egreedy_actor()
      • make_ppo_ac()
      • make_ppo_actor()
    • gomoku_rl.policy.dqn module
      • DQN
      • get_replay_buffer()
    • gomoku_rl.policy.ppo module
      • PPO
  • Module contents
    • get_policy()
    • get_pretrained_policy()
• gomoku_rl.runner package
  • Submodules
    • gomoku_rl.runner.base module
      • Runner
      • SPRunner
    • gomoku_rl.runner.independent_rl_runner module
      • IndependentRLRunner
      • IndependentRLSPRunner
    • gomoku_rl.runner.psro_runner module
      • PSRORunner
      • PSROSPRunner
  • Module contents
• gomoku_rl.utils package
  • Submodules
    • gomoku_rl.utils.augment module
      • AntiDiagonalFlip
      • DiagonalFlip
      • HorizontalFlip
      • Identity
      • Rotation
      • Transform
      • VerticalFlip
      • augment_transition()
      • get_augmented_transition()
    • gomoku_rl.utils.elo module
      • Elo
      • compute_elo_ratings()
      • compute_expected_score()
    • gomoku_rl.utils.eval module
      • eval_win_rate()
      • get_payoff_matrix()
    • gomoku_rl.utils.log module
      • Mean
      • get_log_func()
    • gomoku_rl.utils.misc module
      • add_prefix()
      • get_kwargs()
      • set_seed()
    • gomoku_rl.utils.module module
      • ActorNet
      • MyDuelingCnnDQNet
      • PolicyHead
      • ResidualBlock
      • ResidualTower
      • ValueHead
      • ValueNet
      • count_parameters()
      • init_params()
    • gomoku_rl.utils.policy module
      • uniform_policy()
    • gomoku_rl.utils.psro module
      • ConvergedIndicator
      • PSROPolicyWrapper
      • PayoffType
      • Population
      • calculate_jpc()
      • get_meta_solver()
      • get_new_payoffs()
      • get_new_payoffs_sp()
      • init_payoffs_sp()
      • print_payoffs()
      • solve_last_n()
      • solve_nash()
      • solve_uniform()
    • gomoku_rl.utils.sampler module
      • SequentialSampler
    • gomoku_rl.utils.test module
      • Type
      • assert_layer_transition()
      • assert_observation()
      • assert_tensor_1d_all()
      • assert_transition()
      • no_nan_in_tensordict()
    • gomoku_rl.utils.visual module
      • annotate_heatmap()
      • heatmap()
      • payoff_headmap()
    • gomoku_rl.utils.wandb module
      • dict_flatten()
      • init_wandb()
  • Module contents

Submodules

• gomoku_rl.collector module
  • BlackPlayCollector
    • BlackPlayCollector.__init__()
    • BlackPlayCollector.reset()
    • BlackPlayCollector.rollout()
  • Collector
    • Collector.reset()
    • Collector.rollout()
  • SelfPlayCollector
    • SelfPlayCollector.__init__()
    • SelfPlayCollector.reset()
    • SelfPlayCollector.rollout()
  • VersusPlayCollector
    • VersusPlayCollector.__init__()
    • VersusPlayCollector.reset()
    • VersusPlayCollector.rollout()
  • WhitePlayCollector
    • WhitePlayCollector.__init__()
    • WhitePlayCollector.reset()
    • WhitePlayCollector.rollout()
  • make_transition()
  • round()
  • self_play_step()
• gomoku_rl.core module
  • Gomoku
    • Gomoku.__init__()
    • Gomoku.get_action_mask()
    • Gomoku.get_encoded_board()
    • Gomoku.is_valid()
    • Gomoku.reset()
    • Gomoku.step()
    • Gomoku.to()
  • compute_done()
• gomoku_rl.env module
  • GomokuEnv
    • GomokuEnv.__init__()
    • GomokuEnv.batch_size
    • GomokuEnv.board_size
    • GomokuEnv.device
    • GomokuEnv.num_envs
    • GomokuEnv.reset()
    • GomokuEnv.set_post_step()
    • GomokuEnv.step()
    • GomokuEnv.step_and_maybe_reset()

Module contents