| { | |
| "type": "transformer_diffusion", | |
| "n_obs_steps": 2, | |
| "input_features": { | |
| "observation.state": { | |
| "type": "STATE", | |
| "shape": [ | |
| 7 | |
| ] | |
| }, | |
| "observation.images.gripper": { | |
| "type": "VISUAL", | |
| "shape": [ | |
| 3, | |
| 400, | |
| 640 | |
| ] | |
| }, | |
| "observation.images.depth": { | |
| "type": "VISUAL", | |
| "shape": [ | |
| 3, | |
| 400, | |
| 640 | |
| ] | |
| } | |
| }, | |
| "output_features": { | |
| "action": { | |
| "type": "ACTION", | |
| "shape": [ | |
| 7 | |
| ] | |
| } | |
| }, | |
| "device": "cuda", | |
| "use_amp": false, | |
| "push_to_hub": true, | |
| "repo_id": null, | |
| "private": null, | |
| "tags": null, | |
| "license": null, | |
| "pretrained_path": null, | |
| "horizon": 12, | |
| "n_action_steps": 8, | |
| "normalization_mapping": { | |
| "VISUAL": "MEAN_STD", | |
| "STATE": "MIN_MAX", | |
| "ACTION": "MIN_MAX" | |
| }, | |
| "vision_backbone": "resnet34", | |
| "pretrained_backbone_weights": "ResNet34_Weights.IMAGENET1K_V1", | |
| "state_dim": 7, | |
| "action_dim": 7, | |
| "d_model": 512, | |
| "nhead": 8, | |
| "num_encoder_layers": 6, | |
| "dim_feedforward": 1024, | |
| "diffusion_step_embed_dim": 128, | |
| "down_dims": [ | |
| 512, | |
| 1024, | |
| 2048 | |
| ], | |
| "kernel_size": 5, | |
| "n_groups": 8, | |
| "use_film_scale_modulation": true, | |
| "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 | |
| } |