MCPcopy Create free account

hub / github.com/RLinf/LaWAM / functions

Functions1,214 in github.com/RLinf/LaWAM

⨍ Functions 1,214◇ Types & classes 176↳ Endpoints 1
↓ 2 callersMethod_normalize_robotwin_eef_quaternions_
( actions: np.ndarray, passthrough_indices: tuple[int, ...] = (), )
examples/Robotwin/starvla_policy/deploy_policy.py:401
↓ 2 callersMethod_normalize_robotwin_eef_quaternions_
( actions: np.ndarray, passthrough_indices: tuple[int, ...] = (), )
examples/Robotwin/eval_files/model2robotwin_interface.py:542
↓ 2 callersFunction_normalize_scheduler_state
(payload: Any)
examples/Robotwin/eval_files/batched_eval_runner.py:228
↓ 2 callersFunction_parse_optional_float
(value: Any)
examples/Robotwin/starvla_policy/deploy_policy.py:168
↓ 2 callersMethod_pop_slot_action
(self, slot_id: int)
examples/Robotwin/eval_files/model2robotwin_interface.py:371
↓ 2 callersMethod_prepare_example
(self, payload: dict[str, Any], images: tuple[np.ndarray, np.ndarray, np.ndarray])
deployment/model_server/server_policy_tcp_fold_towel.py:382
↓ 2 callersMethod_prepare_example
(self, payload: dict[str, Any])
deployment/model_server/server_policy_http_franka.py:457
↓ 2 callersFunction_prepare_output_dir
(output_dir: str | None)
examples/LIBERO/eval_files/libero_eval_core.py:197
↓ 2 callersFunction_prepare_run
(args: argparse.Namespace)
examples/Robotwin/eval_files/batched_eval_runner.py:832
↓ 2 callersMethod_prepare_state_condition
( self, *, state: torch.Tensor, state_mask: torch.Tensor, embodiment_i
starVLA/model/framework/vlas/flowmatching_expert.py:327
↓ 2 callersFunction_prepend_unique_tasks
(tasks: list[str], prefix_tasks: list[str])
examples/Robotwin/eval_files/batched_eval_runner.py:284
↓ 2 callersFunction_reset_runtime_state
()
examples/LIBERO/eval_files/libero_eval_core.py:862
↓ 2 callersFunction_resize_frames_preserve_aspect
( frames: np.ndarray, resize_size: tuple[int, int] | int | None, )
starVLA/dataloader/gr00t_lerobot/video.py:84
↓ 2 callersFunction_resolve_llm_layers_container
(llm_module)
starVLA/model/framework/vlas/vlm_auto.py:218
↓ 2 callersMethod_resolve_slot_key
(self, slot_key: int = 0, kwargs: Optional[dict[str, Any]] = None)
examples/LIBERO/eval_files/model2libero_interface.py:264
↓ 2 callersMethod_run_shared_encoding_core
( self, *, prepared_batch: Union[LatentWorldPolicyInferBatch, LatentWorldPolicyTrainBa
starVLA/model/framework/vlas/lawam.py:609
↓ 2 callersMethod_safe_quantile_1d
Compute quantiles on at most `max_elems` values to avoid backend size limits.
latent_action_model/core/lam_lightinng.py:221
↓ 2 callersMethod_sample_candidate
( self, index: int, max_video_retries: int, )
starVLA/dataloader/gr00t_lerobot/datasets.py:2615
↓ 2 callersFunction_save_episode_frames
(frame_dir: Path, frames: list[np.ndarray])
examples/LIBERO/eval_files/libero_eval_core.py:358
↓ 2 callersMethod_set_nested_value
Set nested value through dot-separated path
starVLA/training/trainer_utils/config_tracker.py:361
↓ 2 callersMethod_slice_state_action_rows
( column_vals, start: int, end: int, expected_dim: int, )
starVLA/dataloader/gr00t_lerobot/datasets.py:1435
↓ 2 callersFunction_suite_success_rate
(episode_results: list[EpisodeResult])
examples/LIBERO/eval_files/libero_eval_core.py:723
↓ 2 callersFunction_task_id_list
(task_suite: Any, num_tasks_to_run: int)
examples/LIBERO/eval_files/libero_eval_core.py:637
↓ 2 callersFunction_task_init_state_index
(initial_states: list[Any], episode_idx: int)
examples/LIBERO/eval_files/libero_eval_core.py:336
↓ 2 callersMethod_to_client_normalized_actions
(self, raw_actions: np.ndarray)
deployment/model_server/server_policy_dataset_replay.py:200
↓ 2 callersMethod_to_cpu_float_tensor
(value, *, field_name: str, ex_idx: int)
starVLA/model/framework/latent_world/batch_builder.py:242
↓ 2 callersMethod_to_float
(v: Any, default: float = 0.0)
latent_action_model/core/lam_lightinng.py:210
↓ 2 callersFunction_to_float64_1d
(timestamps: list[float] | np.ndarray)
starVLA/dataloader/gr00t_lerobot/video.py:42
↓ 2 callersMethod_to_float64_timestamps
(ts_vals)
starVLA/dataloader/gr00t_lerobot/datasets.py:1415
↓ 2 callersMethod_try_get_sample
(self, index: int)
latent_action_model/data_loader/lerobot_dataset.py:263
↓ 2 callersFunction_uint8_to_unit_float
(tensor: torch.Tensor)
latent_action_model/data_loader/video_aug.py:36
↓ 2 callersMethod_unit_to_negone_posone
(self, images01: torch.Tensor)
latent_action_model/core/vjepa_encoder.py:512
↓ 2 callersMethod_validate_interpolation
(cls, interpolation: str)
starVLA/dataloader/gr00t_lerobot/transform/video.py:261
↓ 2 callersMethodapply_single
(self, data: dict)
starVLA/dataloader/gr00t_lerobot/transform/groottransforms.py:255
↓ 2 callersFunctionapply_training_freeze_policy
(model, cfg)
starVLA/training/trainer_utils/trainer_tools.py:294
↓ 2 callersMethodbuild
(self)
starVLA/model/framework/latent_world/config_builder.py:30
↓ 2 callersFunctionbuild_dataloaders
(cfg)
starVLA/dataloader/__init__.py:161
↓ 2 callersMethodbuild_dataset_statistics
Build dataset statistics for the current dataset keyed by embodiment tag. Only includes statistics for keys that are actually used in
starVLA/dataloader/gr00t_lerobot/datasets.py:1752
↓ 2 callersMethodbuild_example
(self, task_description: str, observation: dict[str, Any])
examples/Robotwin/eval_files/model2robotwin_interface.py:288
↓ 2 callersFunctionbuild_framework
Build a framework model from config. Args: cfg: Config object (OmegaConf / namespace) containing: cfg.framework.name: Id
starVLA/model/framework/__init__.py:35
↓ 2 callersFunctionbuild_latent_world_processor_spec
(*, policy_cfg: Any, vlm_model_id: str)
starVLA/model/framework/latent_world/processor_utils.py:17
↓ 2 callersFunctionbuild_placeholder_masks
( input_ids: torch.Tensor, *, act_queries: int, flow_queries: int, placeholder_id: int, )
starVLA/model/framework/latent_world/batch_utils.py:60
↓ 2 callersFunctionbuild_qwenvl_messages
( *, images: Sequence[Sequence[Any]], wrist_images: Sequence[Sequence[Any] | None], instructio
starVLA/model/framework/latent_world/vlm_adapter.py:63
↓ 2 callersFunctionbuild_time_grid
Build per-sample natural-time grids t_i = i / hz and hz-derived token masks. Returns: t_grid: (B, T) float timestamps in seconds.
starVLA/model/framework/vlas/flowmatching_expert.py:45
↓ 2 callersMethodclose
(self)
deployment/model_server/tools/websocket_policy_client.py:72
↓ 2 callersFunctioncollate_gr00t
(features: List[dict], processor)
starVLA/dataloader/gr00t_lerobot/transform/groottransforms.py:32
↓ 2 callersFunctioncollect_task_completion
( run_dir: str | os.PathLike[str], task_names: list[str], *, expected_num_episodes: int | None
examples/Robotwin/eval_files/robotwin_eval_common.py:351
↓ 2 callersMethodcompute_inter_code_stats
(self)
latent_action_model/core/vq.py:548
↓ 2 callersMethodcompute_overall_statistics
Computes overall statistics from per-task statistics using dataset sample weights. Args: per_task_stats: List of per-tas
starVLA/dataloader/gr00t_lerobot/datasets.py:2829
↓ 2 callersFunctionconfigure_latent_world_processor
( processor: Any, *, placeholder_token: str, )
starVLA/model/framework/latent_world/processor_utils.py:25
↓ 2 callersFunctionensure_runtime_paths
(robotwin_path: str | os.PathLike[str], starvla_root: str | os.PathLike[str])
examples/Robotwin/eval_files/robotwin_eval_common.py:207
↓ 2 callersMethodextract_vision_features
(self, videos: torch.Tensor, *, n: Optional[Any] = -2)
latent_action_model/core/lam_model.py:562
↓ 2 callersFunctionfilter_lam_video_keys
Keep all non-wrist camera keys for LAM. Single non-wrist view selection is handled downstream by mixture sampling policy.
starVLA/dataloader/gr00t_lerobot/data_config_lam.py:42
↓ 2 callersMethodflush
(self)
examples/Robotwin/eval_files/robotwin_batch_bridge.py:61
↓ 2 callersMethodforward
(self, x: torch.Tensor)
starVLA/dataloader/gr00t_lerobot/transform/state_action.py:108
↓ 2 callersFunctionget_benchmark_adapter
(variant_name: str)
examples/LIBERO/eval_files/libero_benchmark_adapters.py:27
↓ 2 callersMethodget_identity_ctx
()
starVLA/training/trainer_utils/overwatch.py:112
↓ 2 callersMethodget_key_meta
Get the metadata for a key in the LeRobot modality metadata. Args: key (str): The key to get the metadata for. Returns:
starVLA/dataloader/gr00t_lerobot/schema.py:121
↓ 2 callersMethodget_modality_metadata
(self, key: str)
starVLA/dataloader/gr00t_lerobot/transform/concat.py:217
↓ 2 callersMethodget_raw_actions
(self, source: Any)
deployment/model_server/dataset_replay_policy_base.py:106
↓ 2 callersMethodget_server_metadata
(self)
deployment/model_server/tools/websocket_policy_client.py:38
↓ 2 callersMethodget_state_action_dims
Get the dimension of a state or action key from the dataset metadata.
starVLA/dataloader/gr00t_lerobot/transform/concat.py:227
↓ 2 callersMethodget_transform
(self, mode: Literal["train", "eval"] = "train")
starVLA/dataloader/gr00t_lerobot/transform/video.py:278
↓ 2 callersFunctiongpu_two_view_video_aug
GPU-side two-view video augmentation. Args: videos_uint8: [B,T,H,W,C] or [B,T,C,H,W] uint8 tensor. Returns: video1, vide
latent_action_model/data_loader/video_aug.py:60
↓ 2 callersFunctionimagenet_normalize_
In-place ImageNet normalization for tensor [...,3,H,W].
latent_action_model/data_loader/video_aug.py:124
↓ 2 callersFunctioninitialize_task_scheduler
(run_dir: str | os.PathLike[str], scheduled_tasks: list[str])
examples/Robotwin/eval_files/batched_eval_runner.py:270
↓ 2 callersFunctioninspect_task_completion
( run_dir: str | os.PathLike[str], task_name: str, *, expected_num_episodes: int | None = None
examples/Robotwin/eval_files/robotwin_eval_common.py:254
↓ 2 callersFunctioninstall_interrupt_handlers
()
examples/Robotwin/eval_files/batched_eval_runner.py:42
↓ 2 callersMethodinverse
(self, x: torch.Tensor)
starVLA/dataloader/gr00t_lerobot/transform/state_action.py:194
↓ 2 callersFunctioninvert_gripper_action
(action: np.ndarray)
examples/LIBERO/eval_files/libero_eval_core.py:149
↓ 2 callersFunctioninvert_gripper_value
(gripper: np.ndarray)
examples/Robotwin/starvla_policy/deploy_policy.py:25
↓ 2 callersFunctioninvert_gripper_value
(gripper: np.ndarray)
examples/Robotwin/eval_files/model2robotwin_interface.py:15
↓ 2 callersMethodis_rotation_key
(self, key: str)
starVLA/dataloader/gr00t_lerobot/transform/concat.py:234
↓ 2 callersMethodkmeans_init
( self, nodes: Tensor, num_iters: Optional[int] = None, max_samples: Optional[
latent_action_model/core/vq.py:308
↓ 2 callersFunctionlog_full_checkpoint_summary
( *, framework_name: str, checkpoint_path: str | Path, checkpoint_keys: set[str], prefix_c
starVLA/model/framework/base_framework.py:68
↓ 2 callersFunctionmake_LeRobotSingleDataset
Make a LeRobotSingleDataset object. :param data_root_dir: The root directory of the dataset. :param data_name: The name of the dataset.
starVLA/dataloader/lerobot_datasets.py:143
↓ 2 callersMethodmodel_dump
(self, *args, **kwargs)
starVLA/dataloader/gr00t_lerobot/transform/state_action.py:228
↓ 2 callersMethodmodel_dump
(self, *args, **kwargs)
starVLA/dataloader/gr00t_lerobot/transform/concat.py:74
↓ 2 callersFunctionmodulate
AdaLN modulation: x * (1 + scale) + shift
latent_action_model/core/utils/lam_decoder.py:8
↓ 2 callersFunctionnorm_cdf
(x)
latent_action_model/core/utils/utils.py:8
↓ 2 callersMethodorthogonality_loss
(self)
latent_action_model/core/vq.py:458
↓ 2 callersMethodplot_usage_distribution
(self, usage, filename)
latent_action_model/core/lam_lightinng.py:876
↓ 2 callersFunctionprecompute_single_dataset
( data_root_dir: Path, dataset_name: str, robot_type: str, data_cfg: dict[str, Any], )
latent_action_model/precompute_dataset_cache.py:117
↓ 2 callersMethodpredict_action
( self, *, batch: LatentWorldPolicyInferBatch, guidance_scale: Optional[float]
starVLA/model/framework/vlas/lawam.py:793
↓ 2 callersFunctionprepare_frame_spatial_uint8
( frame: np.ndarray, target_hw: Tuple[int, int], *, apply_center_crop_90: bool = False, )
starVLA/model/framework/latent_world/batch_utils.py:175
↓ 2 callersFunctionquat2axisangle
(quat: np.ndarray)
examples/LIBERO/eval_files/libero_benchmark_adapters.py:46
↓ 2 callersFunctionrelease_claimed_task
( run_dir: str | os.PathLike[str], *, task_name: str, worker_index: int, expected_num_epis
examples/Robotwin/eval_files/batched_eval_runner.py:373
↓ 2 callersMethodreplace_unused_codebooks
(self)
latent_action_model/core/vq.py:577
↓ 2 callersMethodreset
(self, task_description: Optional[str] = None)
examples/Robotwin/starvla_policy/deploy_policy.py:517
↓ 2 callersFunctionresolve_robotwin_python
()
examples/Robotwin/eval_files/batched_eval_runner.py:79
↓ 2 callersFunctionresolve_starvla_python
()
examples/Robotwin/eval_files/batched_eval_runner.py:67
↓ 2 callersFunctionresolve_tasks_from_env
()
examples/Robotwin/eval_files/robotwin_eval_common.py:141
↓ 2 callersFunctionsafe_hash
(input_tuple)
starVLA/dataloader/gr00t_lerobot/datasets.py:2243
↓ 2 callersMethodsample_time
Sample diffusion time tau aligned with Isaac-GR00T gr00t_n1d6: tau = (1 - Beta(alpha, beta)) * noise_s. Keep this as the sin
starVLA/model/framework/vlas/flowmatching_expert.py:357
↓ 2 callersFunctionselect_incomplete_tasks
( run_dir: str | os.PathLike[str], task_names: list[str], *, expected_num_episodes: int | None
examples/Robotwin/eval_files/robotwin_eval_common.py:367
↓ 2 callersFunctionsend_action_request
(seed: int)
examples/Robotwin/eval_files/robotwin_batch_bridge.py:362
↓ 2 callersMethodseparate_positions
(self, ids, H_patches, W_patches)
latent_action_model/core/utils/modules.py:228
↓ 2 callersMethodseparate_positions
(self, ids, H_patches=None, W_patches=None)
latent_action_model/core/utils/modules.py:399
↓ 2 callersMethodset_flow_train_step
(self, step: int)
starVLA/model/framework/vlas/lawam.py:535
↓ 2 callersMethodset_mixture_epoch
Set epoch on train mixture dataset if available. Returns: bool: True if epoch was applied, False otherwise.
latent_action_model/data_loader/lerobot_datamodule.py:140
← previousnext →201–300 of 1,214, ranked by callers