tf.contrib.model_pruning

 

upper level

 

Modules

 

None

 

Classes

 

MaskedBasicLSTMCell:

Basic LSTM recurrent network cell with pruning.

MaskedLSTMCell:

LSTMCell with pruning.

Pruning:

 

Functions

 

apply_mask(…):

Apply mask to a given weight tensor.

get_masked_weights(…):

get_masks(…):

get_pruning_hparams(…):

Get a tf.HParams object with the default values for the hyperparameters.

get_thresholds(…):

get_weight_sparsity(…):

Get sparsity of the weights.

get_weights(…):

masked_conv2d(…):

Adds an 2D convolution followed by an optional batch_norm layer.

masked_convolution(…):

Adds an 2D convolution followed by an optional batch_norm layer.

masked_fully_connected(…):

Adds a sparse fully connected layer. The weight matrix is masked.

train(…):

Wrapper around tf-slim’s train function.