def get_initial_feedables(self) -> DecoderFeedables:
feedables = AutoregressiveDecoder.get_initial_feedables(self)
rnn_feedables = RNNFeedables(
prev_contexts=[tf.zeros([self.batch_size, a.context_vector_size])
for a in self.attentions],
prev_rnn_state=self.initial_state,
prev_rnn_output=self.initial_state)
return feedables._replace(other=rnn_feedables)
def get_initial_feedables(self) -> DecoderFeedables:
feedables = AutoregressiveDecoder.get_initial_feedables(self)
rnn_feedables = RNNFeedables(prev_contexts=[
tf.zeros([self.batch_size, a.context_vector_size])
for a in self.attentions
],
prev_rnn_state=self.initial_state,
prev_rnn_output=self.initial_state)
return feedables._replace(other=rnn_feedables)
def get_initial_feedables(self) -> DecoderFeedables:
feedables = AutoregressiveDecoder.get_initial_feedables(self)
tr_feedables = TransformerFeedables(
input_sequence=tf.zeros(
shape=[self.batch_size, 0, self.dimension],
dtype=tf.float32,
name="input_sequence"),
input_mask=tf.zeros(
shape=[self.batch_size, 0, 1],
dtype=tf.float32,
name="input_mask"))
return feedables._replace(other=tr_feedables)