{ "type": "dit", "n_obs_steps": 5, "normalization_mapping": { "VISUAL": "MEAN_STD", "STATE": "MIN_MAX", "ACTION": "MIN_MAX" }, "input_features": { "observation.images.image": { "type": "VISUAL", "shape": [ 3, 256, 256 ] }, "observation.images.wrist_image": { "type": "VISUAL", "shape": [ 3, 256, 256 ] }, "observation.state": { "type": "STATE", "shape": [ 8 ] }, "observation.state.joint": { "type": "STATE", "shape": [ 7 ] } }, "output_features": { "action": { "type": "ACTION", "shape": [ 7 ] } }, "device": "cuda", "use_amp": false, "horizon": 16, "n_action_steps": 8, "vision_backbone_size": 18, "pretrained_backbone_weights": null, "avg_pool": true, "vision_backbone_norm_name": "group_norm", "vision_backbone_norm_num_groups": 16, "use_spatial_softmax": false, "vision_backbone": "resnet18", "crop_shape": null, "crop_is_random": false, "use_group_norm": true, "spatial_softmax_num_keypoints": 32, "dim_model": 512, "n_heads": 8, "dropout": 0.1, "dim_feedforward": 2048, "feedforward_activation": "gelu", "n_encoder_layers": 6, "n_decoder_layers": 6, "noise_scheduler_type": "DDIM", "num_train_timesteps": 100, "beta_schedule": "squaredcos_cap_v2", "beta_start": 0.0001, "beta_end": 0.02, "prediction_type": "epsilon", "clip_sample": true, "set_alpha_to_one": true, "steps_offset": 0, "num_inference_steps": 10, "do_mask_loss_for_padding": false, "optimizer_lr": 0.0001, "optimizer_betas": [ 0.95, 0.999 ], "optimizer_eps": 1e-08, "optimizer_weight_decay": 1e-06, "optimizer_grad_clip_norm": 10.0, "scheduler_name": "cosine", "scheduler_warmup_steps": 500 }