behaviors:
  Pyramids:
    behavioral_cloning: null
    checkpoint_interval: 500000
    hyperparameters: {batch_size: 128, beta: 0.01, beta_schedule: linear, buffer_size: 2048,
      epsilon: 0.2, epsilon_schedule: linear, lambd: 0.95, learning_rate: 0.0003,
      learning_rate_schedule: linear, num_epoch: 3}
    init_path: null
    keep_checkpoints: 5
    max_steps: 1000000
    network_settings: {deterministic: false, goal_conditioning_type: hyper, hidden_units: 512,
      memory: null, normalize: false, num_layers: 2, vis_encode_type: simple}
    reward_signals:
      extrinsic:
        gamma: 0.99
        network_settings: {deterministic: false, goal_conditioning_type: hyper, hidden_units: 128,
          memory: null, normalize: false, num_layers: 2, vis_encode_type: simple}
        strength: 1.0
      rnd:
        encoding_size: null
        gamma: 0.99
        learning_rate: 0.0001
        network_settings: {deterministic: false, goal_conditioning_type: hyper, hidden_units: 64,
          memory: null, normalize: false, num_layers: 3, vis_encode_type: simple}
        strength: 0.01
    self_play: null
    summary_freq: 30000
    threaded: false
    time_horizon: 128
    trainer_type: ppo
checkpoint_settings: {force: false, inference: false, initialize_from: null, load_model: false,
  results_dir: results, resume: false, run_id: Pyramids Training, train_model: false}
debug: false
default_settings: null
engine_settings: {capture_frame_rate: 60, height: 84, no_graphics: true, quality_level: 5,
  target_frame_rate: -1, time_scale: 20, width: 84}
env_settings: {base_port: 5005, env_args: null, env_path: ./trained-envs-executables/linux/Pyramids/Pyramids,
  max_lifetime_restarts: 10, num_areas: 1, num_envs: 1, restarts_rate_limit_n: 1,
  restarts_rate_limit_period_s: 60, seed: -1}
environment_parameters: null
torch_settings: {device: null}