config.yaml

resume: false
device: cuda
use_amp: false
seed: 1000
dataset_repo_id: iantc104/gv_sim_hook_package_2arms
video_backend: pyav
training:
  offline_steps: 15625
  num_workers: 4
  batch_size: 16
  eval_freq: 1000
  log_freq: 100
  save_checkpoint: true
  save_freq: 1000
  online_steps: 0
  online_rollout_n_episodes: 1
  online_rollout_batch_size: 1
  online_steps_between_rollouts: 1
  online_sampling_ratio: 0.5
  online_env_seed: null
  online_buffer_capacity: null
  online_buffer_seed_size: 0
  do_online_rollout_async: false
  image_transforms:
    enable: false
    max_num_transforms: 3
    random_order: false
    brightness:
      weight: 1
      min_max:
      - 0.8
      - 1.2
    contrast:
      weight: 1
      min_max:
      - 0.8
      - 1.2
    saturation:
      weight: 1
      min_max:
      - 0.5
      - 1.5
    hue:
      weight: 1
      min_max:
      - -0.05
      - 0.05
    sharpness:
      weight: 1
      min_max:
      - 0.8
      - 1.2
  lr: 2.5e-05
  lr_backbone: 1.0e-05
  weight_decay: 0.0001
  grad_clip_norm: 10
  delta_timestamps:
    action:
    - 0.0
    - 0.04
    - 0.08
    - 0.12
    - 0.16
    - 0.2
    - 0.24
    - 0.28
    - 0.32
    - 0.36
    - 0.4
    - 0.44
    - 0.48
    - 0.52
    - 0.56
    - 0.6
    - 0.64
    - 0.68
    - 0.72
    - 0.76
    - 0.8
    - 0.84
    - 0.88
    - 0.92
    - 0.96
    - 1.0
    - 1.04
    - 1.08
    - 1.12
    - 1.16
    - 1.2
    - 1.24
    - 1.28
    - 1.32
    - 1.36
    - 1.4
    - 1.44
    - 1.48
    - 1.52
    - 1.56
    - 1.6
    - 1.64
    - 1.68
    - 1.72
    - 1.76
    - 1.8
    - 1.84
    - 1.88
    - 1.92
    - 1.96
eval:
  n_episodes: 1
  batch_size: 1
  use_async_envs: false
wandb:
  enable: true
  disable_artifact: false
  project: lerobot
  notes: ''
fps: 25
env:
  name: guided_vision
  task: HookPackage-2Arms-v0
  state_dim: 14
  action_dim: 14
  fps: ${fps}
  episode_length: 300
override_dataset_stats:
  observation.images.overhead_cam:
    mean:
    - - - 0.485
    - - - 0.456
    - - - 0.406
    std:
    - - - 0.229
    - - - 0.224
    - - - 0.225
  observation.images.worms_eye_cam:
    mean:
    - - - 0.485
    - - - 0.456
    - - - 0.406
    std:
    - - - 0.229
    - - - 0.224
    - - - 0.225
  observation.images.wrist_cam_left:
    mean:
    - - - 0.485
    - - - 0.456
    - - - 0.406
    std:
    - - - 0.229
    - - - 0.224
    - - - 0.225
  observation.images.wrist_cam_right:
    mean:
    - - - 0.485
    - - - 0.456
    - - - 0.406
    std:
    - - - 0.229
    - - - 0.224
    - - - 0.225
policy:
  name: act
  n_obs_steps: 1
  chunk_size: 50
  n_action_steps: 50
  input_shapes:
    observation.images.overhead_cam:
    - 3
    - 480
    - 640
    observation.images.worms_eye_cam:
    - 3
    - 480
    - 640
    observation.images.wrist_cam_left:
    - 3
    - 480
    - 640
    observation.images.wrist_cam_right:
    - 3
    - 480
    - 640
    observation.state:
    - ${env.state_dim}
  output_shapes:
    action:
    - ${env.action_dim}
  input_normalization_modes:
    observation.images.overhead_cam: mean_std
    observation.images.worms_eye_cam: mean_std
    observation.images.wrist_cam_left: mean_std
    observation.images.wrist_cam_right: mean_std
    observation.state: mean_std
  output_normalization_modes:
    action: mean_std
  vision_backbone: resnet18
  pretrained_backbone_weights: ResNet18_Weights.IMAGENET1K_V1
  replace_final_stride_with_dilation: false
  pre_norm: false
  dim_model: 512
  n_heads: 8
  dim_feedforward: 3200
  feedforward_activation: relu
  n_encoder_layers: 4
  n_decoder_layers: 1
  use_vae: true
  latent_dim: 32
  n_vae_encoder_layers: 4
  temporal_ensemble_coeff: null
  dropout: 0.1
  kl_weight: 10.0