Trains with the given dataset. .. code-block:: python algo.fit(episodes, n_steps=1000000) Args: dataset: li

Append observation, action, reward and terminal flag to buffer. If the terminal flag is True, Monte-Carlo returns will be computed with

Returns algorithm configured with json file. The Json file should be the one saved during fitting. .. code-block:: python

Returns scorer function of evaluation on environment. This function returns scorer function, which is suitable to the standard scikit-learn s

Checks value and returns EncoderFactory object. Returns: d3rlpy.encoders.EncoderFactory: encoder factory object.

Returns PyTorch's state enocder module. Args: observation_shape: observation shape. Returns: an enocder obje

Returns stacked observation. Returns: numpy.ndarray: stacked observation.

Returns registered Q function factory object. Args: name: registered Q function factory type name. kwargs: Q function arguments.

Checks value and returns Device object. Returns: d3rlpy.gpu.Device: device object.

Instantiate implementation objects with the dataset shapes. This method will be used internally when `fit` method is called. Args:

Checks value and returns QFunctionFactory object. Returns: d3rlpy.q_functions.QFunctionFactory: Q function factory object.

Start training loop of online deep reinforcement learning. Args: env: gym-like environment. buffer : replay buffer.

Returns sampled mini-batch of transitions. If observation is image, you can stack arbitrary frames via ``n_frames``. .. code

Stack new image. Args: image (numpy.ndarray): image observation.

Returns GPU id. Returns: GPU id.

Returns the number of appended elements in buffer. Returns: the number of elements in buffer.

Clear stacked observation by filling 0.

Returns processed observations. Args: x: observation. Returns: processed observation.

Returns PyTorch's state-action enocder module. Args: observation_shape: observation shape. action_size: action size.

Returns a scaler type. Returns: scaler type.

Returns atari dataset and envrironment. The dataset is provided through d4rl-atari. See more details including available dataset from its Git

Returns cartpole dataset and environment. The dataset is automatically downloaded to ``d3rlpy_data/cartpole.h5`` if it does not exist. A

Registers reward scaler class. Args: cls: scaler class inheriting ``RewardScaler``.

Returns reversely transformed observations. Args: x: observation. Returns: reversely transformed observation

r"""Returns Soft Off-Policy Classification metrics. This function returns scorer function, which is suitable to the standard scikit-learn sco

Returns decayed :math:`\\epsilon`. Returns: :math:`\\epsilon`.