results/ppo/config.yaml

num_threads: 1
float32_matmul_precision: high
dry_run: false
seed: 42
torch_use_deterministic_algorithms: false
torch_backends_cudnn_benchmark: true
torch_backends_cudnn_deterministic: false
cublas_workspace_config: null
exp_name: ppo_doapp
run_name: 2024-04-15_15-25-55_ppo_doapp_42
root_dir: ppo/doapp
algo:
  name: ppo
  total_steps: 1024
  per_rank_batch_size: 16
  run_test: true
  cnn_keys:
    encoder:
    - frame
  mlp_keys:
    encoder:
    - own_character
    - own_health
    - own_side
    - own_wins
    - opp_character
    - opp_health
    - opp_side
    - opp_wins
    - stage
    - timer
    - action
  optimizer:
    _target_: torch.optim.Adam
    lr: 0.005
    eps: 1.0e-06
    weight_decay: 0
    betas:
    - 0.9
    - 0.999
  anneal_lr: false
  gamma: 0.99
  gae_lambda: 0.95
  update_epochs: 1
  loss_reduction: mean
  normalize_advantages: true
  clip_coef: 0.2
  anneal_clip_coef: false
  clip_vloss: false
  ent_coef: 0.0
  anneal_ent_coef: false
  vf_coef: 1.0
  rollout_steps: 32
  dense_units: 16
  mlp_layers: 1
  dense_act: torch.nn.Tanh
  layer_norm: false
  max_grad_norm: 1.0
  encoder:
    cnn_features_dim: 128
    mlp_features_dim: 32
    dense_units: 16
    mlp_layers: 1
    dense_act: torch.nn.Tanh
    layer_norm: false
  actor:
    dense_units: 16
    mlp_layers: 1
    dense_act: torch.nn.Tanh
    layer_norm: false
  critic:
    dense_units: 16
    mlp_layers: 1
    dense_act: torch.nn.Tanh
    layer_norm: false
buffer:
  size: 32
  memmap: true
  validate_args: false
  from_numpy: false
  share_data: false
checkpoint:
  every: 100
  resume_from: null
  save_last: true
  keep_last: 5
distribution:
  validate_args: false
env:
  id: doapp
  num_envs: 1
  frame_stack: 1
  sync_env: true
  screen_size: 64
  action_repeat: 1
  grayscale: false
  clip_rewards: false
  capture_video: true
  frame_stack_dilation: 1
  max_episode_steps: null
  reward_as_observation: false
  wrapper:
    _target_: sheeprl.envs.diambra.DiambraWrapper
    id: doapp
    action_space: DISCRETE
    screen_size: 64
    grayscale: false
    repeat_action: 1
    rank: null
    log_level: 0
    increase_performance: true
    diambra_settings:
      role: P1
      step_ratio: 6
      difficulty: 4
      continue_game: 0.0
      show_final: false
      outfits: 2
      splash_screen: false
    diambra_wrappers:
      stack_actions: 1
      no_op_max: 0
      no_attack_buttons_combinations: false
      add_last_action: true
      scale: false
      exclude_image_scaling: false
      process_discrete_binary: false
      role_relative: true
fabric:
  _target_: lightning.fabric.Fabric
  devices: 1
  num_nodes: 1
  strategy: auto
  accelerator: cpu
  precision: 32-true
  callbacks:
  - _target_: sheeprl.utils.callback.CheckpointCallback
    keep_last: 5
metric:
  log_every: 5000
  disable_timer: false
  log_level: 1
  sync_on_compute: false
  aggregator:
    _target_: sheeprl.utils.metric.MetricAggregator
    raise_on_missing: false
    metrics:
      Rewards/rew_avg:
        _target_: torchmetrics.MeanMetric
        sync_on_compute: false
      Game/ep_len_avg:
        _target_: torchmetrics.MeanMetric
        sync_on_compute: false
  logger:
    _target_: lightning.fabric.loggers.TensorBoardLogger
    name: 2024-04-15_15-25-55_ppo_doapp_42
    root_dir: logs/runs/ppo/doapp
    version: null
    default_hp_metric: true
    prefix: ''
    sub_dir: null
model_manager:
  disabled: true
  models: {}