MCPcopy Create free account
hub / github.com/chinawithfrank/ChatBotCourse / seq2seq_f

Method seq2seq_f

chatbotv4/seq2seq_model.py:133–144  ·  view source on GitHub ↗
(encoder_inputs, decoder_inputs, do_decode)

Source from the content-addressed store, hash-verified

131
132 # The seq2seq function: we use embedding for the input and attention.
133 def seq2seq_f(encoder_inputs, decoder_inputs, do_decode):
134 return seq2seq_patch.embedding_attention_seq2seq(
135 encoder_inputs,
136 decoder_inputs,
137 cell,
138 cell,
139 num_encoder_symbols=source_vocab_size,
140 num_decoder_symbols=target_vocab_size,
141 embedding_size=size,
142 output_projection=output_projection,
143 feed_previous=do_decode,
144 dtype=dtype)
145
146 # Feeds for inputs.
147 self.encoder_inputs = []

Callers

nothing calls this directly

Calls

no outgoing calls

Tested by

no test coverage detected