{ "type": "drifting", "n_obs_steps": 1, "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 ] } }, "output_features": { "action": { "type": "ACTION", "shape": [ 8 ] } }, "device": "cuda", "use_amp": false, "use_peft": false, "push_to_hub": true, "repo_id": "Xihe666/drifting_robomme", "private": null, "tags": null, "license": null, "pretrained_path": null, "pretrained_revision": null, "chunk_size": 40, "n_action_steps": 40, "max_state_dim": 132, "max_action_dim": 132, "normalization_mapping": { "VISUAL": "IDENTITY", "STATE": "IDENTITY", "ACTION": "IDENTITY" }, "base_model_path": "nvidia/GR00T-N1.7-3B", "action_decode_transform": null, "embodiment_tag": "new_embodiment", "tune_llm": false, "tune_visual": false, "tune_projector": true, "tune_diffusion_model": false, "tune_vlln": true, "tune_top_llm_layers": 0, "num_inference_timesteps": 1, "rtc_ramp_rate": null, "use_flash_attention": false, "use_relative_actions": false, "relative_exclude_joints": [], "optimizer_lr": 0.0001, "optimizer_betas": [ 0.9, 0.999 ], "optimizer_eps": 1e-08, "optimizer_weight_decay": 1e-05, "warmup_ratio": 0.05, "use_bf16": true, "model_params_fp32": true, "image_size": [ 256, 256 ], "tokenizer_assets_repo": null, "lora_rank": 0, "lora_alpha": 16, "lora_dropout": 0.1, "lora_full_model": false, "video_backend": "decord", "balance_dataset_weights": true, "balance_trajectory_weights": true, "dataset_paths": null, "output_dir": "./tmp/gr00t", "save_steps": 1000, "max_steps": 10000, "batch_size": 32, "dataloader_num_workers": 8, "report_to": "wandb", "resume": false, "backbone_model_name": "nvidia/Cosmos-Reason2-2B", "backbone_embedding_dim": 2048, "select_layer": 16, "state_history_length": 1, "max_num_embodiments": 32, "action_model_dim": 1024, "action_model_num_layers": 12, "action_model_num_heads": 16, "action_model_ff_dim": 4096, "action_model_dropout": 0.1, "attend_text_every_n_blocks": 2, "add_pos_embed": true, "max_seq_len": 1024, "num_timestep_buckets": 1000, "proximal_time": 0.9, "proximal_loss_weight": 81.0, "geometry_epsilon": 1e-06, "state_dropout_prob": 0.2, "tune_action_model": true }