MCPcopy Create free account

hub / github.com/GalaxyGeneralRobotics/LATENT / functions

Functions332 in github.com/GalaxyGeneralRobotics/LATENT

⨍ Functions 332◇ Types & classes 35
Methodreset
(self, rng: jax.Array, trajectory_data: TrajectoryData = None)
latent_mj/envs/g1_tracking/train/g1_env_tracking_tennis_dr.py:290
Methodreset
(self)
latent_mj/envs/g1_tracking/play/play_g1_env_tracking_tennis.py:233
Methodreset_state
(self, carry, backend)
latent_mj/utils/dataset/traj_handler.py:27
Methodreset_state
(self, carry, backend=jnp)
latent_mj/utils/dataset/traj_handler.py:284
Methodreset_state_simple
(self, rng)
latent_mj/utils/dataset/traj_handler.py:340
Functionrotate_obs
Function to rotate a state (or set of states) around the rotation axis. Args: state (list or np.array): Single state or multiple sta
latent_mj/utils/math.py:9
Methodrun_evaluation
Run one epoch of evaluation.
latent_mj/learning/policy/acting.py:126
Methodset_transition_len
(self, transition_len)
latent_mj/utils/dataset/start_end_transition_handler.py:488
Functionsgd_step
( carry, unused_t, data: types.Transition, normalizer_params: running_statisti
latent_mj/learning/policy/ppo/train_tracking.py:344
Functionsingle_step
(carry, _)
latent_mj/envs/g1_tracking/train/g1_env_tracking_tennis_dr.py:259
Functionsingle_step
(carry, _)
latent_mj/envs/g1_tracking/train/g1_env_tracking_tennis.py:252
Methodstep
(self, state: State, action: jax.Array, trajectory_data)
latent_mj/envs/g1_tracking/utils/wrapper.py:52
Methodstep
(self, state: mjx_env.State, action: jax.Array, trajectory_data)
latent_mj/envs/g1_tracking/utils/wrapper.py:109
Methodstep
(self, state: State, action: np.ndarray)
latent_mj/envs/g1_tracking/play/play_g1_env_tracking_tennis.py:350
Methodstop
(self)
latent_mj/envs/g1_tracking/train/g1_env_tracking_tennis.py:967
Functiontask_to_xml
(with_racket: bool=False)
latent_mj/envs/g1_tracking/g1_tracking_constants_tennis.py:28
Methodto_dict
(self)
latent_mj/utils/dataset/traj_class.py:202
Methodto_jax
(self)
latent_mj/utils/dataset/traj_class.py:598
Methodto_jax
(self)
latent_mj/utils/dataset/traj_handler.py:481
Methodto_jax_fp32
(self)
latent_mj/utils/dataset/traj_class.py:1069
Methodto_jax_fp8
(self)
latent_mj/utils/dataset/traj_class.py:1060
Methodto_jnp
(self)
latent_mj/utils/dataset/traj_class.py:1290
Methodto_np
(self)
latent_mj/utils/dataset/traj_class.py:1293
Methodto_numpy
(self)
latent_mj/utils/dataset/traj_class.py:592
Functiontrain
PPO training. Args: environment: the environment to train num_timesteps: the total number of environment steps to use during trai
latent_mj/learning/policy/ppo/train_tracking.py:112
Functiontraining_step
( carry: Tuple[TrainingState, envs.State, PRNGKey], unused_t )
latent_mj/learning/policy/ppo/train_tracking.py:367
Methodtraj_render
(self, traj_info: TrajCarry, record: bool = False)
latent_mj/envs/g1_tracking/train/g1_env_tracking_tennis.py:954
Functiontransform_angle_2pi
Transforms an angle to be in [-pi, pi]. Args: angle (float): Angle in radians. Returns: Angle in radians in [-pi, pi].
latent_mj/utils/math.py:37
Methodupdate_episode_metrics
(self, aggregated_metrics, dones, training_metrics)
latent_mj/learning/policy/ppo/metrics_aggregator.py:36
Functionupdate_file_handler
Adds or updates file logging.
latent_mj/utils/logger.py:47
Functionwrap_fn
Common wrapper pattern for all brax training agents. Args: env: environment to be wrapped vision: whether the environment will be vis
latent_mj/envs/g1_tracking/utils/wrapper.py:118
Methodxml_path
(self)
latent_mj/envs/g1_tracking/train/base_env.py:77
← previous301–332 of 332, ranked by callers