cambrian ======== .. py:module:: cambrian .. autoapi-nested-parse:: Cambrian package init file. Submodules ---------- .. toctree:: :maxdepth: 1 /reference/api/cambrian/agents/index /reference/api/cambrian/config/index /reference/api/cambrian/envs/index /reference/api/cambrian/eyes/index /reference/api/cambrian/main/index /reference/api/cambrian/ml/index /reference/api/cambrian/renderer/index /reference/api/cambrian/utils/index Attributes ---------- .. autoapisummary:: cambrian.__author__ cambrian.__license__ Classes ------- .. autoapisummary:: cambrian.MjCambrianConfig cambrian.MjCambrianTrainer Package Contents ---------------- .. py:class:: MjCambrianConfig Bases: :py:obj:`hydra_config.HydraContainerConfig` The base config for the mujoco cambrian environment. Used for type hinting. :ivar logdir: The primary directory which simulation data is stored in. This is the highest level directory used for the experiment. `expdir` is the subdirectory used for a specific experiment. :vartype logdir: Path :ivar expdir: The directory used for a specific experiment. This is the directory where the experiment's data is stored. Should evaluate to `logdir / `expsubdir` :vartype expdir: Path :ivar expsubdir: The subdirectory relative to logdir where the experiment's data is stored. This is the directory where the experiment's data is stored. :vartype expsubdir: Path :ivar expname: The name of the experiment. Used to name the logging subdirectory. :vartype expname: str :ivar seed: The base seed used when initializing the default thread/process. Launched processes should use this seed value to calculate their own seed values. This is used to ensure that each process has a unique seed. :vartype seed: int :ivar training: The config for the training process. :vartype training: MjCambrianTrainingConfig :ivar env: The config for the environment. :vartype env: MjCambrianEnvConfig :ivar eval_env: The config for the evaluation environment. :vartype eval_env: MjCambrianEnvConfig .. py:class:: MjCambrianTrainer(config) This is the trainer class for running training and evaluation. :Parameters: **config** (*MjCambrianConfig*) -- The config to use for training and evaluation. .. py:method:: train() Train the agent. .. py:data:: __author__ :value: 'Camera Culture (a2cc@media.mit.edu)' Camera Culture (a2cc@media.mit.edu) .. py:data:: __license__ :value: 'BSD3' BSD3