| { |
| "dataset": { |
| "select_dataset": "HumanoidStation_in_jushen", |
| "repo_ids": null, |
| "roots": null, |
| "repo_id": null, |
| "root": null, |
| "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 |
| ] |
| } |
| } |
| } |
| }, |
| "local_files_only": false, |
| "use_imagenet_stats": true, |
| "video_backend": "pyav" |
| }, |
| "env": null, |
| "policy": { |
| "type": "crossvla_stage2_plus", |
| "n_obs_steps": 2, |
| "action_sample_factor": 1, |
| "normalization_mapping": { |
| "VISUAL": "IDENTITY", |
| "STATE": "MEAN_STD", |
| "ACTION": "MEAN_STD" |
| }, |
| "input_features": { |
| "observation.state": { |
| "type": "STATE", |
| "shape": [ |
| 32 |
| ] |
| }, |
| "observation.images.image_0": { |
| "type": "VISUAL", |
| "shape": [ |
| 224, |
| 3, |
| 224 |
| ] |
| }, |
| "observation.images.image_1": { |
| "type": "VISUAL", |
| "shape": [ |
| 224, |
| 3, |
| 224 |
| ] |
| }, |
| "observation.images.image_2": { |
| "type": "VISUAL", |
| "shape": [ |
| 224, |
| 3, |
| 224 |
| ] |
| }, |
| "observation.images.image_3": { |
| "type": "VISUAL", |
| "shape": [ |
| 224, |
| 3, |
| 224 |
| ] |
| }, |
| "observation.images.image_wrist_0": { |
| "type": "VISUAL", |
| "shape": [ |
| 224, |
| 3, |
| 224 |
| ] |
| }, |
| "observation.images.image_wrist_1": { |
| "type": "VISUAL", |
| "shape": [ |
| 224, |
| 3, |
| 224 |
| ] |
| }, |
| "observation.state_is_pad": { |
| "type": "STATE", |
| "shape": [ |
| 2 |
| ] |
| }, |
| "observation.images.image_0_is_pad": { |
| "type": "STATE", |
| "shape": [ |
| 2 |
| ] |
| }, |
| "observation.images.image_1_is_pad": { |
| "type": "STATE", |
| "shape": [ |
| 2 |
| ] |
| }, |
| "observation.images.image_2_is_pad": { |
| "type": "STATE", |
| "shape": [ |
| 2 |
| ] |
| }, |
| "observation.images.image_3_is_pad": { |
| "type": "STATE", |
| "shape": [ |
| 2 |
| ] |
| }, |
| "observation.images.image_wrist_0_is_pad": { |
| "type": "STATE", |
| "shape": [ |
| 2 |
| ] |
| }, |
| "observation.images.image_wrist_1_is_pad": { |
| "type": "STATE", |
| "shape": [ |
| 2 |
| ] |
| } |
| }, |
| "output_features": { |
| "action": { |
| "type": "ACTION", |
| "shape": [ |
| 32 |
| ] |
| }, |
| "action_is_pad": { |
| "type": "ACTION", |
| "shape": [ |
| 50 |
| ] |
| } |
| }, |
| "stage1_pretrained_path": "/media/jushen/bamboo-fan/Save/crossvla/stage1_4dataset_10_8_bs12_kl_cb256_0528_stat/checkpoints/275000/pretrained_model/", |
| "stage2_pretrained_path": "None", |
| "stage2_latent_image_token_check": false, |
| "dataset_stats_generate": true, |
| "heterogeneous": true, |
| "split_dataset": true, |
| "real_robot_dev": false, |
| "image_interval_steps": 50, |
| "action_latent_token_num": 13, |
| "mformer_hidden_size": 768, |
| "decoder_hidden_size": 768, |
| "codebook_embed_dim": 256, |
| "codebook_k_size": 256, |
| "action_chunk_size": 50, |
| "chunk_size": 50, |
| "n_action_steps": 50, |
| "resampler": true, |
| "resampler_dim": 2048, |
| "resampler_depth": 3, |
| "resampler_dim_head": 128, |
| "resampler_heads": 4, |
| "resampler_num_media_embeds": 1, |
| "resampler_num_latents": 9, |
| "max_state_dim": 32, |
| "max_action_dim": 32, |
| "resize_imgs_with_padding": [ |
| 224, |
| 224 |
| ], |
| "empty_cameras": 0, |
| "adapt_to_pi_aloha": false, |
| "use_delta_joint_actions_aloha": false, |
| "tokenizer_max_length": 48, |
| "proj_width": 1024, |
| "num_steps": 10, |
| "use_cache": true, |
| "attention_implementation": "eager", |
| "freeze_vision_encoder": false, |
| "freeze_language_encoder": true, |
| "train_expert_only": false, |
| "train_state_proj": true, |
| "optimizer_lr": 0.0001, |
| "optimizer_betas": [ |
| 0.9, |
| 0.95 |
| ], |
| "optimizer_eps": 1e-08, |
| "optimizer_weight_decay": 1e-10, |
| "scheduler_warmup_steps": 5000, |
| "scheduler_decay_steps": 300000, |
| "scheduler_decay_lr": 1e-06 |
| }, |
| "output_dir": "/media/jushen/bamboo-fan/Save/crossvla/dev/stage2_dev_pretrain_10_8_bs8_stage1_275k", |
| "job_name": "stage2_dev_pretrain_10_8_bs8_stage1_275k", |
| "resume": true, |
| "device": "cuda", |
| "use_amp": false, |
| "seed": 1000, |
| "num_workers": 4, |
| "batch_size": 8, |
| "gradient_accumulation_steps": 1, |
| "steps": 300000, |
| "eval_freq": 20000, |
| "log_freq": 200, |
| "save_checkpoint": true, |
| "save_freq": 10000, |
| "use_policy_training_preset": true, |
| "optimizer": { |
| "type": "adamw", |
| "lr": 0.0001, |
| "weight_decay": 1e-10, |
| "grad_clip_norm": 10.0, |
| "betas": [ |
| 0.9, |
| 0.95 |
| ], |
| "eps": 1e-08 |
| }, |
| "scheduler": { |
| "type": "cosine_decay_with_warmup", |
| "num_warmup_steps": 5000, |
| "num_decay_steps": 300000, |
| "peak_lr": 0.0001, |
| "decay_lr": 1e-06 |
| }, |
| "eval": { |
| "n_episodes": 50, |
| "batch_size": 50, |
| "use_async_envs": false |
| }, |
| "wandb": { |
| "enable": true, |
| "disable_artifact": true, |
| "project": "crossvla_dev", |
| "entity": null, |
| "notes": null |
| } |
| } |
