def add_pos_signals(x, hparams, name="pos_emb"):
with tf.variable_scope(name, reuse=False):
if hparams.pos == "timing":
x = common_attention.add_timing_signal_nd(x)
else:
assert hparams.pos == "emb"
x = common_attention.add_positional_embedding_nd(
x, hparams.max_length, name=name)
return x
def add_pos_signals(x, hparams, name="pos_emb"):
with tf.variable_scope(name, reuse=False):
if hparams.pos == "timing":
x = common_attention.add_timing_signal_nd(x)
else:
assert hparams.pos == "emb"
x = common_attention.add_positional_embedding_nd(
x, hparams.max_length, name)
return x
def testAddPositionalEmbeddingNd(self, input_shape):
x = np.random.rand(*input_shape)
y = common_attention.add_positional_embedding_nd(
tf.constant(x, dtype=tf.float32),
max_length=5,
name="pos_embedding")
self.evaluate(tf.global_variables_initializer())
res = self.evaluate(y)
self.assertEqual(res.shape, input_shape)
