:param inputs: (batch, step_length, size, channels=1) :return:

:param inputs: (batch, nb_steps, input_size) :return:

:param inputs: (batch, nb_steps, size) :param initial_state: None or: if directional: (b

:param inputs: if not bidirectional: (batch, size) or [(batch, size)] else:

:param inputs: (batch, size) :return:

:param inputs: (alloutput): (batch, nb_instruments, nb_steps, step_size, input_size, channels) :return:

For the file check_dataset.py :param args: :return:

if batch dim is not in the shape, then put it :param final_shape: tuple<int>: (4,) or (5,): (a, b, c, d) or (?, a, b, c, d)

:param args: :return:

:return:

For the file compute_data.py :param args: :return:

:param replicate: :param save_images: :param save_truth: :param verbose: :param array_truth: :param

:param nb_steps_to_1: :param input_shape: real input of the convolution :param model_param_enc: :param strides: :return:

:param input_shape: (None, (None?), in_features) :return: (None, (None?), out_features)

:param input_shape: List(2)[ (batch, modalities, size) ] :return: List(2)[

:param input_shape: List(2)[ (batch, modalities, nb_steps?, size) ] + List[1](batch, mod

:param input_shape: List(2)[ (batch, modalities, nb_steps?, size) ] + List[1](batch, mod

:param input_shape: List(2)[(batch, nb_instruments, nb_steps, size)] + List(1)[(batch, nb_instrument, nb_steps)] :return:

:param input_shape: (batch, nb_steps, input_size) :return:

:param inputs_shape: [(batch, nb_units), (batch, nb_units), (batch, input_size)] :return:

:param inputs_shape: (batch, size) :return:

:param input_shape: :param model_param_conv: :param strides: :return:

:param loss_options: :param input_param: { nb_instruments, input_size

:param input_param: :param model_param: :param nb_steps: :param optimizer: :return: the neural network

:param loss_options: :param input_param: { nb_instruments, input_size

:return:

:return: Return the dictionary of not changeable parameters: d[name] = value

:param x: return x, if x is str, return eval(x) :param env: dic of environment variables :return: