| { | |
| "type": "groot", | |
| "n_obs_steps": 1, | |
| "input_features": { | |
| "observation.state": { | |
| "type": "STATE", | |
| "shape": [ | |
| 6 | |
| ] | |
| }, | |
| "observation.images.wrist": { | |
| "type": "VISUAL", | |
| "shape": [ | |
| 3, | |
| 480, | |
| 640 | |
| ] | |
| } | |
| }, | |
| "output_features": { | |
| "action": { | |
| "type": "ACTION", | |
| "shape": [ | |
| 6 | |
| ] | |
| } | |
| }, | |
| "device": "cuda", | |
| "use_amp": false, | |
| "push_to_hub": false, | |
| "repo_id": null, | |
| "private": null, | |
| "tags": null, | |
| "license": null, | |
| "pretrained_path": "nvidia/GR00T-N1.5-3B", | |
| "base_model_path": "nvidia/GR00T-N1.5-3B", | |
| "eagle_path": null, | |
| "tokenizer_assets_repo": "lerobot/eagle2hg-processor-groot-n1p5", | |
| "tune_llm": true, | |
| "tune_visual": true, | |
| "eagle_select_layer": 12, | |
| "eagle_project_to_dim": null, | |
| "tune_projector": true, | |
| "tune_diffusion_model": true, | |
| "chunk_size": 16, | |
| "n_action_steps": 16, | |
| "max_state_dim": 64, | |
| "max_action_dim": 32, | |
| "noise_beta_alpha": 1.5, | |
| "noise_beta_beta": 1.0, | |
| "noise_s": 0.999, | |
| "num_timestep_buckets": 1000, | |
| "num_inference_steps": 4, | |
| "max_num_embodiments": 32, | |
| "embodiment_tag": "new_embodiment", | |
| "action_head_hidden_size": 1024, | |
| "action_head_input_embedding_dim": 1536, | |
| "action_head_backbone_embedding_dim": 2048, | |
| "add_pos_embed": true, | |
| "max_seq_len": 1024, | |
| "num_target_vision_tokens": 32, | |
| "use_vlln": true, | |
| "diffusion_model_cfg": { | |
| "attention_head_dim": 48, | |
| "cross_attention_dim": 2048, | |
| "dropout": 0.2, | |
| "final_dropout": true, | |
| "interleave_self_attention": true, | |
| "norm_type": "ada_norm", | |
| "num_attention_heads": 32, | |
| "num_layers": 16, | |
| "output_dim": 1024, | |
| "positional_embeddings": null | |
| }, | |
| "vl_self_attention_cfg": { | |
| "attention_head_dim": 64, | |
| "dropout": 0.2, | |
| "final_dropout": true, | |
| "num_attention_heads": 32, | |
| "num_layers": 4, | |
| "positional_embeddings": null | |
| }, | |
| "image_size": [ | |
| 448, | |
| 448 | |
| ], | |
| "tokenizer_max_length": 200, | |
| "use_bf16": true, | |
| "compile_model": false, | |
| "compile_mode": "max-autotune", | |
| "dtype": "bfloat16", | |
| "normalization_mapping": { | |
| "VISUAL": "IDENTITY", | |
| "STATE": "MEAN_STD", | |
| "ACTION": "MEAN_STD" | |
| }, | |
| "optimizer_lr": 0.0001, | |
| "optimizer_betas": [ | |
| 0.95, | |
| 0.999 | |
| ], | |
| "optimizer_eps": 1e-08, | |
| "optimizer_weight_decay": 1e-05, | |
| "optimizer_grad_clip_norm": 1.0, | |
| "scheduler_warmup_steps": 500, | |
| "scheduler_decay_steps": 10000, | |
| "scheduler_decay_lr": 1e-05 | |
| } |