maximellerbach HF Staff
Drop the bbox image size step: bbox targets are [0, 1] image fractions
bef7b76 verified | { | |
| "name": "policy_preprocessor", | |
| "steps": [ | |
| { | |
| "registry_name": "render_runtime_messages_processor", | |
| "config": { | |
| "recipe": { | |
| "messages": [ | |
| { | |
| "role": "user", | |
| "content": "${task}", | |
| "stream": "low_level", | |
| "target": false, | |
| "if_present": null, | |
| "tool_calls_from": null | |
| }, | |
| { | |
| "role": "assistant", | |
| "content": "BBoxJSON: ${bbox}", | |
| "stream": "low_level", | |
| "target": true, | |
| "if_present": "bbox", | |
| "tool_calls_from": null | |
| }, | |
| { | |
| "role": "assistant", | |
| "content": "Subtask: ${subtask}", | |
| "stream": "low_level", | |
| "target": true, | |
| "if_present": "subtask", | |
| "tool_calls_from": null | |
| } | |
| ], | |
| "bindings": { | |
| "bbox": "emitted_at(t, style=vqa, role=assistant, camera=observation.images.exterior)" | |
| }, | |
| "blend": null, | |
| "weight": null, | |
| "route": null | |
| } | |
| } | |
| }, | |
| { | |
| "registry_name": "render_training_messages_processor", | |
| "config": { | |
| "recipe": null | |
| } | |
| }, | |
| { | |
| "registry_name": "rename_observations_processor", | |
| "config": { | |
| "rename_map": {} | |
| } | |
| }, | |
| { | |
| "registry_name": "to_batch_processor", | |
| "config": {} | |
| }, | |
| { | |
| "registry_name": "g05_image_transform", | |
| "config": { | |
| "camera_order": [ | |
| "observation.images.head_camera", | |
| "observation.images.left_camera", | |
| "observation.images.right_camera" | |
| ], | |
| "camera_sizes": { | |
| "observation.images.head_camera": [ | |
| 256, | |
| 256 | |
| ], | |
| "observation.images.left_camera": [ | |
| 256, | |
| 256 | |
| ], | |
| "observation.images.right_camera": [ | |
| 256, | |
| 256 | |
| ] | |
| }, | |
| "mean": [ | |
| 0.5, | |
| 0.5, | |
| 0.5 | |
| ], | |
| "std": [ | |
| 0.5, | |
| 0.5, | |
| 0.5 | |
| ], | |
| "optional_camera_keys": [] | |
| } | |
| }, | |
| { | |
| "registry_name": "g05_relative_joint_actions", | |
| "config": { | |
| "enabled": false, | |
| "exclude_joints": [], | |
| "action_names": null, | |
| "num_obs_steps": 1 | |
| } | |
| }, | |
| { | |
| "registry_name": "g05_embodiment_projection", | |
| "config": { | |
| "embodiment": "robotwin", | |
| "policy_state_dim": 20, | |
| "policy_action_dim": 20, | |
| "camera_order": [ | |
| "observation.images.head_camera", | |
| "observation.images.left_camera", | |
| "observation.images.right_camera" | |
| ] | |
| } | |
| }, | |
| { | |
| "registry_name": "normalizer_processor", | |
| "config": { | |
| "eps": 1e-08, | |
| "features": { | |
| "observation.state": { | |
| "type": "STATE", | |
| "shape": [ | |
| 20 | |
| ] | |
| }, | |
| "observation.images.head_camera": { | |
| "type": "VISUAL", | |
| "shape": [ | |
| 3, | |
| 256, | |
| 256 | |
| ] | |
| }, | |
| "observation.images.left_camera": { | |
| "type": "VISUAL", | |
| "shape": [ | |
| 3, | |
| 256, | |
| 256 | |
| ] | |
| }, | |
| "observation.images.right_camera": { | |
| "type": "VISUAL", | |
| "shape": [ | |
| 3, | |
| 256, | |
| 256 | |
| ] | |
| }, | |
| "action": { | |
| "type": "ACTION", | |
| "shape": [ | |
| 20 | |
| ] | |
| } | |
| }, | |
| "norm_map": { | |
| "STATE": "QUANTILES", | |
| "ACTION": "QUANTILES", | |
| "VISUAL": "IDENTITY" | |
| } | |
| }, | |
| "state_file": "policy_preprocessor_step_5_normalizer_processor.safetensors" | |
| }, | |
| { | |
| "registry_name": "g05_normalization_clamp", | |
| "config": { | |
| "minimum": -5.0, | |
| "maximum": 5.0 | |
| } | |
| }, | |
| { | |
| "registry_name": "device_processor", | |
| "config": { | |
| "device": "cuda", | |
| "float_dtype": null | |
| } | |
| }, | |
| { | |
| "registry_name": "g05_tokenizer", | |
| "config": { | |
| "processor_dir": "hf_processor", | |
| "action_tokenizer_path": "action_tokenizer.safetensors" | |
| }, | |
| "artifacts": { | |
| "processor_dir": "hf_processor", | |
| "action_tokenizer_path": "action_tokenizer.safetensors" | |
| } | |
| } | |
| ] | |
| } |