config.json

{
    "type": "diffusion",
    "n_obs_steps": 2,
    "normalization_mapping": {
        "STATE": "MIN_MAX",
        "ENV": "MIN_MAX",
        "ACTION": "MIN_MAX",
        "VISUAL": "IDENTITY"
    },
    "input_features": {
        "observation.state": {
            "type": "STATE",
            "shape": [
                2
            ]
        },
        "observation.environment_state": {
            "type": "ENV",
            "shape": [
                16
            ]
        }
    },
    "output_features": {
        "action": {
            "type": "ACTION",
            "shape": [
                2
            ]
        }
    },
    "device": "cpu",
    "use_amp": false,
    "horizon": 16,
    "n_action_steps": 8,
    "drop_n_last_frames": 7,
    "vision_backbone": "resnet18",
    "crop_shape": [
        84,
        84
    ],
    "crop_is_random": true,
    "pretrained_backbone_weights": null,
    "use_group_norm": true,
    "spatial_softmax_num_keypoints": 32,
    "use_separate_rgb_encoder_per_camera": false,
    "state_backbone": "MLP",
    "state_encoder_block_channels": [
        64,
        128
    ],
    "state_encoder_feature_dim": 256,
    "state_encoder_use_layernorm": true,
    "down_dims": [
        512,
        1024,
        2048
    ],
    "kernel_size": 5,
    "n_groups": 8,
    "diffusion_step_embed_dim": 128,
    "use_film_scale_modulation": true,
    "noise_scheduler_type": "DDPM",
    "num_train_timesteps": 100,
    "beta_schedule": "squaredcos_cap_v2",
    "beta_start": 0.0001,
    "beta_end": 0.02,
    "prediction_type": "epsilon",
    "clip_sample": true,
    "clip_sample_range": 1.0,
    "variance_type": "fixed_small",
    "num_inference_steps": null,
    "do_mask_loss_for_padding": false,
    "optimizer_type": "adamw",
    "optimizer_lr": 0.0001,
    "optimizer_betas": [
        0.95,
        0.999
    ],
    "optimizer_eps": 1e-08,
    "optimizer_weight_decay": 1e-06,
    "scheduler_name": "cosine",
    "scheduler_warmup_steps": 500
}