Return the path to a binary, or `None` if it's not on the path. Returns: A bytestring.

:param item: :return: x + [mask], y x: List(nb_instruments)[(batch, nb_steps, step_size, input_size, channels)]

:param sequence: KerasSequence :param batch_size:

:param sequence: Sequence to copy (Keras Sequence) :param nb_steps_per_file: :param batch_size: Attributs:

:type predict_offset: :param path: The path to the data :param nb_steps: The number of steps in the inputs :param wo

:param out_features: :param bias: :param indep_weights:

:param filters_list: List[List[int]]: :param dropout: float: :param time_stride: int: :type last_pool: bool:

:param filters_list: List[List[int]]: :param dropout: float: :param time_stride: int: :param shapes_after_upsize: Op

:param decoder_param: { 'conv': [[4, 8], [16, 8], [4, 8]], 'dense': [24, 12] } :param dropout_c: flo

:param axis: (not batch in it)

:param axis: (not batch in it) : Axis of the modalities

:param axis: (not batch in it)

:param weight: :param sum_axis: To sum through time. If is None, no summation is done :param kwargs:

:param args: :param kwargs:

:param axis: Batch axis is not included

:param axis: :param axis_reversed:

:param units: int :param dropout:

:param size_list: list<int>, (nb_blocks,) :param dropout: float

:param args: :param activation_layer: :param kwargs:

:param filters_list: List[List[int]]: :param dropout: float: :param time_stride: int: :type last_pool: bool:

:type last_activation: object :param filters_list: List[List[int]]: :param dropout: float: :param time_stride: int:

:param decoder_param: { 'conv': [[4, 8], [16, 8], [4, 8]], 'dense': [24, 12] } :param dropout_c: flo

:param softmax_axis: int:

:param softmax_axis: int:

:param softmax_axis: :param names:

:param size: :param args: :param dropout: :param return_sequences: :param bidirectional: :param kwar

:param state_as_output: :param size_list: :param dropout: :param return_sequences: If True, returns the all sequence

:param nb_steps: :param dropout: :param bidirectional: :param args: :param kwargs:

:param size_list: List[int] :param nb_steps: :param dropout: :param args: :param dropout: :param bid

:param filters: int: :param strides: Tuple[int]: :param dropout: float: :param final_shape: Optional[Tuple[int]]

:param filters: int: the size of the filters :param strides: tuple<int>: (2,): :param dropout: float:

:param filters: int: :param strides: Tuple[int]: :param dropout: float: :param final_shape: Optional[Tuple[int]]

:param args: :param l_semitone: :param l_tone: :param l_tritone: :param mono: :param kwargs:

:param layers: a list of Keras layer :param args: :param kwargs:

:param layers: a list of Keras layer :param args: :param kwargs:

:param weight: the K.variable :param start_value: the value to start :param final_value: the value to have at the end of the

:param filepath:

:param mono:

:param args: :param argtype: :param kwargs:

:type plot_pianoroll: Boolean :param model: The model to generate the notes :param instrument: The instrument the player wan

:param name: The name of the model :param work_on: either 'beat' or 'measure' :param data: if not None, load the data

:param notes_range: note range 0 <= x <= 88 0 : A 87 : C