| { |
| "dataset": { |
| "type": "WSA_Large", |
| "repo_id": "multidata_from_file", |
| "repo_id_file": "outputs/WSA_Large/_repo_id_files/WSA_Large-wsa_large-multidata-delta-pretrain-2026_05_07_15_59_20.txt", |
| "root": null, |
| "episodes": null, |
| "image_transforms": { |
| "enable": false, |
| "preset": null, |
| "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, |
| "use_external_stats": true, |
| "external_stats_path": null, |
| "external_stats_root": "norm_stats_32", |
| "weight_rules_path": "configs/weight_rules_wsa_large_pretrain.yaml", |
| "video_backend": "pyav", |
| "streaming": false, |
| "dist_loading": true, |
| "buffer_size": 1024, |
| "action_mode": "delta", |
| "repack_transforms": { |
| "inputs": [], |
| "outputs": [] |
| }, |
| "data_transforms": { |
| "inputs": [], |
| "outputs": [] |
| }, |
| "model_transforms": { |
| "inputs": [], |
| "outputs": [] |
| }, |
| "dataset_dirs": [], |
| "image_keys": [ |
| "image0", |
| "image1", |
| "image2" |
| ], |
| "image_raw_shapes": [ |
| [ |
| 3, |
| 224, |
| 224 |
| ], |
| [ |
| 3, |
| 224, |
| 224 |
| ], |
| [ |
| 3, |
| 224, |
| 224 |
| ] |
| ], |
| "image_shapes": [ |
| [ |
| 3, |
| 224, |
| 224 |
| ], |
| [ |
| 3, |
| 224, |
| 224 |
| ], |
| [ |
| 3, |
| 224, |
| 224 |
| ] |
| ], |
| "action_keys": [ |
| "default" |
| ], |
| "action_raw_shapes": [ |
| 24 |
| ], |
| "action_shapes": [ |
| 24 |
| ], |
| "state_keys": [ |
| "default" |
| ], |
| "state_raw_shapes": [ |
| 24 |
| ], |
| "state_shapes": [ |
| 24 |
| ], |
| "num_frames": 33, |
| "global_sample_stride": 1, |
| "action_video_freq_ratio": 4, |
| "video_size": [ |
| 384, |
| 320 |
| ], |
| "standardize_video_size_by_cameras": true, |
| "camera_key": null, |
| "val_set_proportion": 0.0, |
| "skip_padding_as_possible": false, |
| "max_padding_retry": 3, |
| "concat_multi_camera": "robotwin", |
| "override_instruction": null, |
| "return_future_3d_images": true, |
| "future_3d_target_index": -1, |
| "processor_num_output_cameras": 3, |
| "processor_action_output_dim": 24, |
| "processor_proprio_output_dim": 24, |
| "processor_resize_shape": null, |
| "processor_use_stepwise_action_norm": false, |
| "processor_norm_default_mode": "z-score", |
| "processor_use_zh_instruction": false, |
| "processor_delta_action_dim_mask": { |
| "default": [ |
| true, |
| true, |
| true, |
| true, |
| true, |
| true, |
| false |
| ] |
| }, |
| "pretrain_multi_embodiment": true, |
| "text_embedding_cache_dir": "outputs/WSA_Large/text_embeds/pretrain", |
| "context_len": 128, |
| "normalization_stats_path": null, |
| "dataset_sampling_weights": [] |
| }, |
| "env": null, |
| "policy": { |
| "type": "WSA_Large", |
| "n_obs_steps": 1, |
| "input_features": { |
| "observation.state": { |
| "type": "STATE", |
| "shape": [ |
| 24 |
| ] |
| } |
| }, |
| "output_features": { |
| "action": { |
| "type": "ACTION", |
| "shape": [ |
| 24 |
| ] |
| } |
| }, |
| "device": "cuda:0", |
| "use_amp": false, |
| "push_to_hub": false, |
| "repo_id": "zaleni/WSA-Large", |
| "private": false, |
| "tags": null, |
| "license": null, |
| "pretrained_path": null, |
| "variant": "wsa_large", |
| "model_id": "Wan-AI/Wan2.2-TI2V-5B", |
| "tokenizer_model_id": "Wan-AI/Wan2.1-T2V-1.3B", |
| "tokenizer_max_len": 128, |
| "load_text_encoder": false, |
| "redirect_common_files": true, |
| "mot_checkpoint_mixed_attn": false, |
| "skip_dit_load_from_pretrain": false, |
| "action_dit_pretrained_path": "checkpoints/wsa_large/ActionDiT_linear_interp_Wan22_alphascale_1024hdim.pt", |
| "future_3d_pretrained_path": "checkpoints/wsa_large/Future3DExpert_linear_interp_Wan22_alphascale_768hdim.pt", |
| "native_checkpoint_path": null, |
| "dtype": "bfloat16", |
| "action_dim": 24, |
| "proprio_dim": 24, |
| "action_horizon": 32, |
| "n_action_steps": 24, |
| "num_inference_steps": 10, |
| "video_dit_config": { |
| "has_image_input": false, |
| "patch_size": [ |
| 1, |
| 2, |
| 2 |
| ], |
| "in_dim": 48, |
| "hidden_dim": 3072, |
| "ffn_dim": 14336, |
| "freq_dim": 256, |
| "text_dim": 4096, |
| "out_dim": 48, |
| "num_heads": 24, |
| "attn_head_dim": 128, |
| "num_layers": 30, |
| "eps": 1e-06, |
| "seperated_timestep": true, |
| "require_clip_embedding": false, |
| "require_vae_embedding": false, |
| "fuse_vae_embedding_in_latents": true, |
| "use_gradient_checkpointing": false, |
| "video_attention_mask_mode": "first_frame_causal", |
| "action_conditioned": false, |
| "action_dim": 24, |
| "action_group_causal_mask_mode": "group_diagonal" |
| }, |
| "action_dit_config": { |
| "action_dim": 24, |
| "hidden_dim": 1024, |
| "ffn_dim": 4096, |
| "num_heads": 24, |
| "attn_head_dim": 128, |
| "num_layers": 30, |
| "text_dim": 4096, |
| "freq_dim": 256, |
| "eps": 1e-06, |
| "use_gradient_checkpointing": false |
| }, |
| "future_3d_config": { |
| "hidden_dim": 768, |
| "ffn_dim": 3072, |
| "text_dim": 4096, |
| "freq_dim": 256, |
| "eps": 1e-06, |
| "num_heads": 24, |
| "attn_head_dim": 128, |
| "num_layers": 30, |
| "num_query_tokens": 432, |
| "da3_num_views": 3, |
| "da3_tokens_per_view": 1296, |
| "da3_query_dim": 2048, |
| "query_layer_indices": [ |
| 14, |
| 19, |
| 24, |
| 29 |
| ], |
| "future_query_init_std": 0.02, |
| "query_mode": "slot_noise", |
| "query_noise_scale": 0.5, |
| "query_noise_min_sigma": 0.0, |
| "query_noise_max_sigma": 0.5, |
| "query_sigma_source": "constant", |
| "slot_pos_scale": 0.5, |
| "use_gradient_checkpointing": false |
| }, |
| "lambda_3d": 0.1, |
| "da3_variant": "large", |
| "da3_teacher_layers": [ |
| 11, |
| 15, |
| 19, |
| 23 |
| ], |
| "da3_query_dim": 2048, |
| "da3_tokens_per_view": 1296, |
| "da3_num_views": 3, |
| "future_3d_tokens_per_view": 144, |
| "future_3d_view_attention_layout": "robotwin", |
| "da3_model_path_or_name": "depth-anything/DA3-LARGE-1.1", |
| "da3_model_name": null, |
| "da3_code_root": null, |
| "da3_teacher_process_res": 504, |
| "da3_layer_weights": [ |
| 1.0, |
| 1.2, |
| 1.4, |
| 1.6 |
| ], |
| "future_query_init_std": 0.02, |
| "future_3d_query_mode": "slot_noise", |
| "future_3d_query_noise_scale": 0.5, |
| "future_3d_query_noise_min_sigma": 0.0, |
| "future_3d_query_noise_max_sigma": 0.5, |
| "future_3d_query_sigma_source": "constant", |
| "future_3d_slot_pos_scale": 0.5, |
| "future_3d_target_index": -1, |
| "log_da3_teacher_timing": false, |
| "video_scheduler": { |
| "train_shift": 5.0, |
| "infer_shift": 5.0, |
| "num_train_timesteps": 1000 |
| }, |
| "action_scheduler": { |
| "train_shift": 5.0, |
| "infer_shift": 5.0, |
| "num_train_timesteps": 1000 |
| }, |
| "lambda_video": 1.0, |
| "lambda_action": 1.0, |
| "action_norm_use_stepwise": false, |
| "action_norm_default_mode": "z-score", |
| "action_norm_exception_mode": null, |
| "action_stats_path": null, |
| "optimizer_lr": 0.0001, |
| "optimizer_betas": [ |
| 0.9, |
| 0.95 |
| ], |
| "optimizer_eps": 1e-08, |
| "optimizer_weight_decay": 0.01, |
| "optimizer_grad_clip_norm": 1.0, |
| "scheduler_type": "cosine", |
| "scheduler_warmup_steps": 5000, |
| "scheduler_warmup_ratio": 0.05, |
| "scheduler_decay_lr": 1e-05, |
| "train_num_epochs": null, |
| "train_max_steps": 300000 |
| }, |
| "output_dir": "outputs/WSA_Large/WSA_Large-wsa_large-multidata-delta-pretrain-2026_05_07_15_59_20", |
| "job_name": "WSA_Large-wsa_large-multidata-delta-pretrain-2026_05_07_15_59_20", |
| "resume": false, |
| "seed": 42, |
| "num_workers": 14, |
| "batch_size": 12, |
| "gradient_accumulation_steps": 1, |
| "steps": 300000, |
| "eval_freq": 20000, |
| "log_freq": 100, |
| "save_checkpoint": true, |
| "save_freq": 20000, |
| "use_policy_training_preset": true, |
| "optimizer": { |
| "type": "adamw", |
| "lr": 0.0001, |
| "weight_decay": 0.01, |
| "grad_clip_norm": 1.0, |
| "betas": [ |
| 0.9, |
| 0.95 |
| ], |
| "eps": 1e-08 |
| }, |
| "scheduler": { |
| "type": "wsa_large_native", |
| "num_warmup_steps": 5000, |
| "peak_lr": 0.0001, |
| "min_lr_ratio": 0.1, |
| "scheduler_type": "cosine", |
| "warmup_ratio": 0.05 |
| }, |
| "eval": { |
| "n_episodes": 50, |
| "batch_size": 50, |
| "use_async_envs": false |
| }, |
| "wandb": { |
| "enable": false, |
| "disable_artifact": true, |
| "project": "WSA_Large", |
| "entity": null, |
| "notes": null, |
| "run_id": null, |
| "mode": null |
| }, |
| "checkpoint_path": null, |
| "rename_map": {} |
| } |
