Module: tf.contrib.seq2seq

Module tf.contrib.seq2seq

Ops for building neural network seq2seq decoders and losses.

Members

attention_decoder_fn_inference(...): Attentional decoder function for dynamic_rnn_decoder during inference.

attention_decoder_fn_train(...): Attentional decoder function for dynamic_rnn_decoder during training.

dynamic_rnn_decoder(...): Dynamic RNN decoder for a sequence-to-sequence model specified by

prepare_attention(...): Prepare keys/values/functions for attention.

sequence_loss(...): Weighted cross-entropy loss for a sequence of logits (per example).

simple_decoder_fn_inference(...): Simple decoder function for a sequence-to-sequence model used in the

simple_decoder_fn_train(...): Simple decoder function for a sequence-to-sequence model used in the

Defined in tensorflow/contrib/seq2seq/__init__.py.