caffe2.python.models.seq2seq.seq2seq_util.LSTMWithAttentionDecoder Class Reference

Inheritance diagram for caffe2.python.models.seq2seq.seq2seq_util.LSTMWithAttentionDecoder:

Public Member Functions
def	scope (self, name)
def	__init__ (self, encoder_outputs, encoder_output_dim, encoder_lengths, vocab_size, attention_type, embedding_size, decoder_num_units, decoder_cells, residual_output_layers=None, name=None, weighted_encoder_outputs=None)
def	get_state_names (self)
def	get_outputs_with_grads (self)
def	get_output_dim (self)
def	get_attention_weights (self)
def	apply (self, model, input_t, seq_lengths, states, timestep)
def	apply_over_sequence (self, model, inputs, seq_lengths, initial_states)

Public Attributes
	name
	num_layers
	cell
	use_attention
	decoder_output_dim
	output_indices

Detailed Description

Definition at line 320 of file seq2seq_util.py.

---

The documentation for this class was generated from the following file:

• caffe2/python/models/seq2seq/seq2seq_util.py