def __init__(self, data_hparams, hparams=None):
ModelBase.__init__(self, hparams)
self._data_hparams = HParams(data_hparams,
PairedTextData.default_hparams())
self._src_vocab = None
self._tgt_vocab = None
self._src_embedder = None
self._tgt_embedder = None
self._connector = None
self._encoder = None
self._decoder = None
def _input_fn():
data = PairedTextData(hparams)
iterator = data.dataset.make_initializable_iterator()
tf.add_to_collection(tf.GraphKeys.TABLE_INITIALIZERS,
iterator.initializer)
batch = iterator.get_next()
features, labels = {}, {}
for key, value in batch.items():
if key.startswith('source_'):
features[key] = value
else:
labels[key] = value
return features, labels
def _build_vocab(self):
self._src_vocab, self._tgt_vocab = PairedTextData.make_vocab(
self._data_hparams.source_dataset,
self._data_hparams.target_dataset)