"""GaussianCNNRegressorModel."""
import numpy as np
import tensorflow as tf
from garage.tf.models import GaussianCNNModel
[docs]class GaussianCNNRegressorModel(GaussianCNNModel):
"""GaussianCNNRegressor based on garage.tf.models.Model class.
This class can be used to perform regression by fitting a Gaussian
distribution to the outputs.
Args:
input_shape(tuple[int]): Input shape of the model (without the batch
dimension).
filters (Tuple[Tuple[int, Tuple[int, int]], ...]): Number and dimension
of filters. For example, ((3, (3, 5)), (32, (3, 3))) means there
are two convolutional layers. The filter for the first layer have 3
channels and its shape is (3 x 5), while the filter for the second
layer have 32 channels and its shape is (3 x 3).
strides(tuple[int]): The stride of the sliding window. For example,
(1, 2) means there are two convolutional layers. The stride of the
filter for first layer is 1 and that of the second layer is 2.
padding (str): The type of padding algorithm to use,
either 'SAME' or 'VALID'.
output_dim (int): Output dimension of the model.
name (str): Model name, also the variable scope.
hidden_sizes (list[int]): Output dimension of dense layer(s) for
the Convolutional model for mean. For example, (32, 32) means the
network consists of two dense layers, each with 32 hidden units.
hidden_nonlinearity (callable): Activation function for intermediate
dense layer(s). It should return a tf.Tensor. Set it to
None to maintain a linear activation.
hidden_w_init (callable): Initializer function for the weight
of intermediate dense layer(s). The function should return a
tf.Tensor.
hidden_b_init (callable): Initializer function for the bias
of intermediate dense layer(s). The function should return a
tf.Tensor.
output_nonlinearity (callable): Activation function for output dense
layer. It should return a tf.Tensor. Set it to None to
maintain a linear activation.
output_w_init (callable): Initializer function for the weight
of output dense layer(s). The function should return a
tf.Tensor.
output_b_init (callable): Initializer function for the bias
of output dense layer(s). The function should return a
tf.Tensor.
learn_std (bool): Is std trainable.
init_std (float): Initial value for std.
adaptive_std (bool): Is std a neural network. If False, it will be a
parameter.
std_share_network (bool): Boolean for whether mean and std share
the same network.
std_filters (Tuple[Tuple[int, Tuple[int, int]], ...]): Number and
dimension of filters. For example, ((3, (3, 5)), (32, (3, 3)))
means there are two convolutional layers. The filter for the first
layer have 3 channels and its shape is (3 x 5), while the filter
for the second layer have 32 channels and its shape is (3 x 3).
std_strides(tuple[int]): The stride of the sliding window. For example,
(1, 2) means there are two convolutional layers. The stride of the
filter for first layer is 1 and that of the second layer is 2.
std_padding (str): The type of padding algorithm to use in std network,
either 'SAME' or 'VALID'.
std_hidden_sizes (list[int]): Output dimension of dense layer(s) for
the Conv for std. For example, (32, 32) means the Conv consists
of two hidden layers, each with 32 hidden units.
min_std (float): If not None, the std is at least the value of min_std,
to avoid numerical issues.
max_std (float): If not None, the std is at most the value of max_std,
to avoid numerical issues.
std_hidden_nonlinearity (callable): Nonlinearity for each hidden layer
in the std network.
std_hidden_w_init (callable): Initializer function for the weight
of intermediate dense layer(s) in the std network.
std_hidden_b_init (callable): Initializer function for the bias
of intermediate dense layer(s) in the std network.
std_output_nonlinearity (callable): Activation function for output
dense layer in the std network. It should return a tf.Tensor. Set
it to None to maintain a linear activation.
std_output_w_init (callable): Initializer function for the weight
of output dense layer(s) in the std network.
std_parameterization (str): How the std should be parametrized. There
are two options:
- exp: the logarithm of the std will be stored, and applied a
exponential transformation
- softplus: the std will be computed as log(1+exp(x))
layer_normalization (bool): Bool for using layer normalization or not.
"""
def __init__(self,
input_shape,
output_dim,
filters,
strides,
padding,
hidden_sizes,
name='GaussianCNNRegressorModel',
hidden_nonlinearity=tf.nn.tanh,
hidden_w_init=tf.initializers.glorot_uniform(),
hidden_b_init=tf.zeros_initializer(),
output_nonlinearity=None,
output_w_init=tf.initializers.glorot_uniform(),
output_b_init=tf.zeros_initializer(),
learn_std=True,
adaptive_std=False,
std_share_network=False,
init_std=1.0,
min_std=1e-6,
max_std=None,
std_filters=(),
std_strides=(),
std_padding='SAME',
std_hidden_sizes=(32, 32),
std_hidden_nonlinearity=tf.nn.tanh,
std_hidden_w_init=tf.initializers.glorot_uniform(),
std_hidden_b_init=tf.zeros_initializer(),
std_output_nonlinearity=None,
std_output_w_init=tf.initializers.glorot_uniform(),
std_parameterization='exp',
layer_normalization=False):
super().__init__(output_dim=output_dim,
filters=filters,
strides=strides,
padding=padding,
hidden_sizes=hidden_sizes,
hidden_nonlinearity=hidden_nonlinearity,
hidden_w_init=hidden_w_init,
hidden_b_init=hidden_b_init,
output_nonlinearity=output_nonlinearity,
output_w_init=output_w_init,
output_b_init=output_b_init,
learn_std=learn_std,
adaptive_std=adaptive_std,
std_share_network=std_share_network,
init_std=init_std,
min_std=min_std,
max_std=max_std,
std_filters=std_filters,
std_strides=std_strides,
std_padding=std_padding,
std_hidden_sizes=std_hidden_sizes,
std_hidden_nonlinearity=std_hidden_nonlinearity,
std_hidden_w_init=std_hidden_w_init,
std_hidden_b_init=std_hidden_b_init,
std_output_nonlinearity=std_output_nonlinearity,
std_output_w_init=std_output_w_init,
std_parameterization=std_parameterization,
layer_normalization=layer_normalization)
self._input_shape = input_shape
[docs] def network_output_spec(self):
"""Network output spec.
Return:
list[str]: List of key(str) for the network outputs.
"""
return [
'sample', 'means', 'log_stds', 'std_param', 'normalized_means',
'normalized_log_stds', 'x_mean', 'x_std', 'y_mean', 'y_std', 'dist'
]
def _build(self, state_input, name=None):
"""Build model given input placeholder(s).
Args:
state_input (tf.Tensor): Place holder for state input.
name (str): Inner model name, also the variable scope of the
inner model, if exist. One example is
garage.tf.models.Sequential.
Return:
tf.Tensor: Sampled action.
tf.Tensor: Mean.
tf.Tensor: Parameterized log_std.
tf.Tensor: log_std.
garage.tf.distributions.DiagonalGaussian: Policy distribution.
"""
with tf.compat.v1.variable_scope('normalized_vars'):
x_mean_var = tf.compat.v1.get_variable(
name='x_mean',
shape=(1, ) + self._input_shape,
dtype=np.float32,
initializer=tf.zeros_initializer(),
trainable=False)
x_std_var = tf.compat.v1.get_variable(
name='x_std_var',
shape=(1, ) + self._input_shape,
dtype=np.float32,
initializer=tf.ones_initializer(),
trainable=False)
y_mean_var = tf.compat.v1.get_variable(
name='y_mean_var',
shape=(1, self._output_dim),
dtype=np.float32,
initializer=tf.zeros_initializer(),
trainable=False)
y_std_var = tf.compat.v1.get_variable(
name='y_std_var',
shape=(1, self._output_dim),
dtype=np.float32,
initializer=tf.ones_initializer(),
trainable=False)
normalized_xs_var = (state_input - x_mean_var) / x_std_var
sample, normalized_mean, normalized_log_std, std_param, dist = super(
)._build(normalized_xs_var)
with tf.name_scope('mean_network'):
means_var = normalized_mean * y_std_var + y_mean_var
with tf.name_scope('std_network'):
log_stds_var = normalized_log_std + tf.math.log(y_std_var)
return (sample, means_var, log_stds_var, std_param, normalized_mean,
normalized_log_std, x_mean_var, x_std_var, y_mean_var,
y_std_var, dist)