Module: tf.contrib.legacy_seq2seq

Module tf.contrib.legacy_seq2seq

Deprecated library for creating sequence-to-sequence models in TensorFlow.

Members

attention_decoder(...): RNN decoder with attention for the sequence-to-sequence model.

basic_rnn_seq2seq(...): Basic RNN sequence-to-sequence model.

embedding_attention_decoder(...): RNN decoder with embedding and attention and a pure-decoding option.

embedding_attention_seq2seq(...): Embedding sequence-to-sequence model with attention.

embedding_rnn_decoder(...): RNN decoder with embedding and a pure-decoding option.

embedding_rnn_seq2seq(...): Embedding RNN sequence-to-sequence model.

embedding_tied_rnn_seq2seq(...): Embedding RNN sequence-to-sequence model with tied (shared) parameters.

model_with_buckets(...): Create a sequence-to-sequence model with support for bucketing.

one2many_rnn_seq2seq(...): One-to-many RNN sequence-to-sequence model (multi-task).

rnn_decoder(...): RNN decoder for the sequence-to-sequence model.

sequence_loss(...): Weighted cross-entropy loss for a sequence of logits, batch-collapsed.

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

tied_rnn_seq2seq(...): RNN sequence-to-sequence model with tied encoder and decoder parameters.

Defined in tensorflow/contrib/legacy_seq2seq/__init__.py.