(self, x, sub_layer_x, num=0)

(self, inputs, encoder=True, reuse=False)

(self, q, k, v)

(self, x, sub_layer_x, num=0)

(self, inputs, logits, decoder_index)

(input_, size)

(self, decoder_emb_inp, encoder_outputs,reuse=False)

(self, decoder_outputs, reuse=False)

(fname)

(self, inputs)

(enc_fname, dec_fname, train=True)

(fname, sorted_vocab)

(self, logits)

(self)

(self)

(self, outputs)

(self, q, k, v)

(self, features, labels)

(self, q, k, v)

()

(self, q, k, v)

(self, output)

(self, output)

(self, qs, ks, vs)

(self, q, k, v)

(self, q, k, v)

(self, decoder_inputs, encoder_outputs)

(self, encoder_inputs=None, decoder_inputs=None)

(self, encoder_inputs)

(self, encoder_emb_inp, reuse=False)

(self)

(in_fname, out_fname, normalize_digits=True)

(keys, rev_vocab=None)

(vocab_fname)

(ids, vocab)

(mode)

(dim, sentence_length, dtype=tf.float32)

()

(vocab, line)

(dataset_fnames)

(sequence)

(sequence)

(self)

(self)

(self, num_layers=8, num_heads=8, linear_key_dim=50,

(self, mode, dtype=tf.float32)

(self, num_heads=1, masked=False, linear_key_dim=50,

(self, w1_dim=200, w2_dim=100, dropout=0.1)

(self, num_layers=8, num_heads=8, linear_key_dim=50,

(labels, predictions)

(labels, predictions, weights=None, metrics_collections=None, up

(self)

(run_config, params)

(values)

(data, buffer_size=10000, batch_size=64, scope="train")

(vocab)

()

(shuffle=True)

(self, mode, features, labels, params)

(questions, answers)

(variables, rev_vocab=None, every_n_iter=100)

(config_name)

(tensor, num_heads, dim)

(tensor)