rnnlib

get_module_device

 get_module_device (model)

source

is_cuda_enabled

 is_cuda_enabled (model)

source

repeat_lstm_state

 repeat_lstm_state (state, batch_size)

source

create_lstm_init_state

 create_lstm_init_state (num_layers, num_directions, hidden_size,
                         init_state_learned=True, device=None)

:param hidden_size: :param init_state_learned: :returns: init_state is a input of lstm cells. _init_state is saved as a parameter of model (such as self._init_state)

source

repeat_lstm_cell_state

 repeat_lstm_cell_state (state, batch_size)

source

create_lstm_cell_init_state

 create_lstm_cell_init_state (hidden_size, init_state_learned=True,
                              device=None)

:param hidden_size: :param init_state_learned: :returns: init_state is a input of lstm cells. _init_state is saved as a parameter of model (such as self._init_state)

source

get_indicator

 get_indicator (length_tensor, max_length=None)

:param length_tensor: :param max_length: :returns: a tensor where positions within ranges are set to 1

source

LSTMCell

 LSTMCell (input_size, hidden_size)

standard LSTM cell

source

LSTMFrame

 LSTMFrame (rnn_cells, batch_first=False, dropout=0, bidirectional=False)

Wrapper of RNNFrame. The ‘for_lstm’ option is always ‘True’.

source

RNNFrame

 RNNFrame (rnn_cells, for_lstm=False, batch_first=False, dropout=0,
           bidirectional=False)

Base class for all neural network modules.

Your models should also subclass this class.

Modules can also contain other Modules, allowing to nest them in a tree structure. You can assign the submodules as regular attributes::

import torch.nn as nn
import torch.nn.functional as F

class Model(nn.Module):
    def __init__(self):
        super().__init__()
        self.conv1 = nn.Conv2d(1, 20, 5)
        self.conv2 = nn.Conv2d(20, 20, 5)

    def forward(self, x):
        x = F.relu(self.conv1(x))
        return F.relu(self.conv2(x))

Submodules assigned in this way will be registered, and will have their parameters converted too when you call :meth:to, etc.

.. note:: As per the example above, an __init__() call to the parent class must be made before assignment on the child.

:ivar training: Boolean represents whether this module is in training or evaluation mode. :vartype training: bool

source

forward_rnn

 forward_rnn (rnn, init_state, input, lengths, batch_first=False,
              embedding:torch.nn.modules.sparse.Embedding=None,
              dropout:torch.nn.modules.dropout.Dropout=None,
              return_packed_output=False)

source

LayerNormLSTM

 LayerNormLSTM (input_size, hidden_size, num_layers=1, batch_first=False,
                dropout=0, r_dropout=0, bidirectional=False,
                layer_norm_enabled=True)

Wrapper of RNNFrame. The ‘for_lstm’ option is always ‘True’.

source

LayerNormLSTMCell

 LayerNormLSTMCell (input_size, hidden_size, dropout=None,
                    layer_norm_enabled=True, cell_ln=None)

It’s based on tf.contrib.rnn.LayerNormBasicLSTMCell Reference: - https://www.tensorflow.org/api_docs/python/tf/contrib/rnn/LayerNormBasicLSTMCell - https://github.com/tensorflow/tensorflow/blob/r1.12/tensorflow/contrib/rnn/python/ops/rnn_cell.py#L1335

source

LayerNormRNNCell

 LayerNormRNNCell (input_size, hidden_size, layer_norm_enabled=True)

Base class for all neural network modules.

Your models should also subclass this class.

Modules can also contain other Modules, allowing to nest them in a tree structure. You can assign the submodules as regular attributes::

import torch.nn as nn
import torch.nn.functional as F

class Model(nn.Module):
    def __init__(self):
        super().__init__()
        self.conv1 = nn.Conv2d(1, 20, 5)
        self.conv2 = nn.Conv2d(20, 20, 5)

    def forward(self, x):
        x = F.relu(self.conv1(x))
        return F.relu(self.conv2(x))

Submodules assigned in this way will be registered, and will have their parameters converted too when you call :meth:to, etc.

.. note:: As per the example above, an __init__() call to the parent class must be made before assignment on the child.

:ivar training: Boolean represents whether this module is in training or evaluation mode. :vartype training: bool