{ "dataset": { "repo_id": "wuc1/bi_so101_ffp_0630", "root": "/mnt/e/dataset/bi_so101_ffp_0623", "episodes": null, "image_transforms": { "enable": false, "max_num_transforms": 3, "random_order": false, "tfs": { "brightness": { "weight": 1.0, "type": "ColorJitter", "kwargs": { "brightness": [ 0.8, 1.2 ] } }, "contrast": { "weight": 1.0, "type": "ColorJitter", "kwargs": { "contrast": [ 0.8, 1.2 ] } }, "saturation": { "weight": 1.0, "type": "ColorJitter", "kwargs": { "saturation": [ 0.5, 1.5 ] } }, "hue": { "weight": 1.0, "type": "ColorJitter", "kwargs": { "hue": [ -0.05, 0.05 ] } }, "sharpness": { "weight": 1.0, "type": "SharpnessJitter", "kwargs": { "sharpness": [ 0.5, 1.5 ] } }, "affine": { "weight": 1.0, "type": "RandomAffine", "kwargs": { "degrees": [ -5.0, 5.0 ], "translate": [ 0.05, 0.05 ] } } } }, "revision": null, "use_imagenet_stats": true, "video_backend": "torchcodec", "return_uint8": false, "streaming": false, "intervention_only": false }, "env": null, "policy": null, "reward_model": { "type": "sarm", "input_features": { "observation.images.left_camera1": { "type": "VISUAL", "shape": [ 480, 640, 3 ] }, "observation.state": { "type": "STATE", "shape": [ 32 ] } }, "output_features": { "sparse_stage": { "type": "REWARD", "shape": [ 13, 1 ] }, "sparse_progress": { "type": "REWARD", "shape": [ 13, 1 ] } }, "device": "cuda", "pretrained_path": null, "push_to_hub": true, "repo_id": "wuc1/sarm_0701", "license": null, "tags": null, "private": null, "annotation_mode": "single_stage", "temporal_window_mode": "bidirectional", "n_obs_steps": 8, "frame_gap": 60, "max_rewind_steps": 4, "image_dim": 512, "text_dim": 512, "hidden_dim": 768, "num_heads": 12, "num_layers": 8, "max_state_dim": 32, "drop_n_last_frames": 1, "batch_size": 64, "clip_batch_size": 64, "dropout": 0.1, "stage_loss_weight": 1.0, "rewind_probability": 0.8, "language_perturbation_probability": 0.2, "num_sparse_stages": 1, "sparse_subtask_names": [ "task" ], "sparse_temporal_proportions": [ 1.0 ], "num_dense_stages": null, "dense_subtask_names": null, "dense_temporal_proportions": null, "pretrained_model_path": null, "image_key": "observation.images.left_camera1", "state_key": "observation.state", "precomputed_image_features_path": "/mnt/e/dataset/bi_so101_ffp_0623/meta/clip_features_left_camera1.npy", "normalization_mapping": { "VISUAL": "IDENTITY", "STATE": "MEAN_STD", "LANGUAGE": "IDENTITY", "REWARD": "IDENTITY" } }, "output_dir": "/home/wy/outputs/train/sarm_0701", "job_name": "sarm_0701", "resume": false, "seed": 1000, "cudnn_deterministic": false, "num_workers": 4, "batch_size": 128, "prefetch_factor": 4, "persistent_workers": true, "steps": 11140, "eval_freq": 20000, "log_freq": 100, "tolerance_s": 0.01, "save_checkpoint": true, "save_freq": 1000, "use_policy_training_preset": true, "optimizer": { "type": "adamw", "lr": 5e-05, "weight_decay": 0.001, "grad_clip_norm": 10.0, "betas": [ 0.9, 0.999 ], "eps": 1e-08 }, "scheduler": { "type": "cosine_decay_with_warmup", "num_warmup_steps": 500, "num_decay_steps": 50000, "peak_lr": 5e-05, "decay_lr": 5e-06 }, "eval": { "n_episodes": 50, "batch_size": 22, "use_async_envs": true }, "validation": { "enable": false, "ratio": 0.2, "freq": 1000, "max_batches": 64, "episodes": null, "split": "tail", "resolved_train_episodes": null, "resolved_val_episodes": null }, "wandb": { "enable": true, "disable_artifact": false, "project": "sarm", "entity": null, "notes": null, "run_id": "7h77mgyj", "mode": null, "add_tags": true }, "peft": null, "use_subtasks": false, "sample_weighting": null, "task_conditioning": null, "rename_map": {}, "checkpoint_path": null }