ADLStream.models.LSTM
Long Short Term Memory (LSTM).
Parameters:
Name | Type | Description | Default |
---|---|---|---|
input_shape |
tuple |
Shape of the input data |
required |
output_size |
int |
Number of neurons of the last layer. |
required |
loss |
tf.keras.Loss |
Loss to be use for training. |
required |
optimizer |
tf.keras.Optimizer |
Optimizer that implements theraining algorithm. |
required |
recurrent_units |
list |
Number of recurrent units for each LSTM layer. Defaults to [64]. |
[64] |
recurrent_dropout |
int between 0 and 1 |
Fraction of the input units to drop. Defaults to 0. |
0 |
return_sequences |
bool |
Whether to return the last output in the output sequence, or the full sequence. Defaults to False. |
False |
dense_layers |
list |
List with the number of hidden neurons for each layer of the dense block before the output. Defaults to []. |
[] |
dense_dropout |
float between 0 and 1 |
Fraction of the dense units to drop. Defaults to 0.0. |
0 |
dense_activation |
tf activation function |
Activation function of the dense layers after the convolutional block. Defaults to "linear". |
'linear' |
out_activation |
tf activation function |
Activation of the output layer. Defaults to "linear". |
'linear' |
Returns:
Type | Description |
---|---|
tf.keras.Model |
LSTM model |
Source code in ADLStream/models/lstm.py
def LSTM(
input_shape,
output_size,
loss,
optimizer,
recurrent_units=[64],
recurrent_dropout=0,
return_sequences=False,
dense_layers=[],
dense_dropout=0,
dense_activation="linear",
out_activation="linear",
):
"""Long Short Term Memory (LSTM).
Args:
input_shape (tuple): Shape of the input data
output_size (int): Number of neurons of the last layer.
loss (tf.keras.Loss): Loss to be use for training.
optimizer (tf.keras.Optimizer): Optimizer that implements theraining algorithm.
recurrent_units (list, optional): Number of recurrent units for each LSTM layer.
Defaults to [64].
recurrent_dropout (int between 0 and 1, optional): Fraction of the input units to drop.
Defaults to 0.
return_sequences (bool, optional): Whether to return the last output in the output sequence, or the full sequence.
Defaults to False.
dense_layers (list, optional): List with the number of hidden neurons for each
layer of the dense block before the output.
Defaults to [].
dense_dropout (float between 0 and 1, optional): Fraction of the dense units to drop.
Defaults to 0.0.
dense_activation (tf activation function, optional): Activation function of the dense
layers after the convolutional block.
Defaults to "linear".
out_activation (tf activation function, optional): Activation of the output layer.
Defaults to "linear".
Returns:
tf.keras.Model: LSTM model
"""
input_shape = input_shape[-len(input_shape) + 1 :]
inputs = tf.keras.layers.Input(shape=input_shape)
x = inputs
if len(input_shape) < 2:
x = tf.keras.layers.Reshape((inputs.shape[1], 1))(x)
# LSTM layers
for i, u in enumerate(recurrent_units):
return_sequences_tmp = (
return_sequences if i == len(recurrent_units) - 1 else True
)
x = tf.keras.layers.LSTM(
u, return_sequences=return_sequences_tmp, dropout=recurrent_dropout
)(x)
# Dense layers
if return_sequences:
x = tf.keras.layers.Flatten()(x)
for hidden_units in dense_layers:
x = tf.keras.layers.Dense(hidden_units, activation=dense_activation)(x)
if dense_dropout > 0:
x = tf.keras.layers.Dropout(dense_dropout)(x)
x = tf.keras.layers.Dense(output_size, activation=out_activation)(x)
model = tf.keras.Model(inputs=inputs, outputs=x)
model.compile(optimizer=optimizer, loss=loss)
return model