def test_string_to_hash_bucket_layer_with_different_ctor_params(self):
# Input space: Batch of strings.
input_space = TextBox(add_batch_rank=True)
# Construct a strong hash bucket with different delimiter, larger number of buckets, string algo and
# int16 dtype.
string_to_hash_bucket = StringToHashBucket(delimiter="-", num_hash_buckets=20, hash_function="strong",
dtype="int16")
test = ComponentTest(component=string_to_hash_bucket, input_spaces=dict(text_inputs=input_space))
# Send a batch of 5 strings through the hash-bucket generator.
inputs = np.array([
"text-A",
"test-B",
"text-C--D-and-E",
"bla bla-D"
])
# NOTE that some different words occupy the same hash bucket (e.g. 'C' and 'and' OR 'text' and [empty]).
# This can be avoided by 1) picking a larger `num_hash_buckets` or 2) using the "strong" hash function.
expected_hash_bucket = np.array([
[2, 6, 18, 18, 18], # text A . . .
[12, 7, 18, 18, 18], # test B . . .
[2, 6, 13, 19, 15], # text C D and E
[13, 13, 18, 18, 18], # bla bla D . . . <- Note that "bla bla" and "D" still have the same bucket (13)
])
expected_lengths = np.array([2, 2, 5, 2])
test.test(("apply", inputs), expected_outputs=(expected_hash_bucket, expected_lengths))
def test_string_to_hash_bucket_layer(self):
# Input space: Batch of strings.
input_space = TextBox(add_batch_rank=True)
# Use a fast-hash function with 10 possible buckets to put a word into.
string_to_hash_bucket = StringToHashBucket(num_hash_buckets=10,
hash_function="fast")
test = ComponentTest(component=string_to_hash_bucket,
input_spaces=dict(text_inputs=input_space))
# Send a batch of 3 strings through the hash-bucket generator.
inputs = np.array(["text A", "test B", "text C D and E"])
# NOTE that some different words occupy the same hash bucket (e.g. 'C' and 'and' (7) OR 'text' and [empty] (3)).
# This can be avoided by 1) picking a larger `num_hash_buckets` or 2) using the "strong" hash function.
expected_hash_bucket = np.array([
[3, 4, 3, 3, 3], # text A . . .
[6, 8, 3, 3, 3], # test B . . .
[3, 7, 5, 7, 2], # text C D and E
])
expected_lengths = np.array([2, 2, 5])
test.test(("call", inputs),
expected_outputs=(expected_hash_bucket, expected_lengths))
def test_functional_api_multi_stream_nn(self):
# Input Space of the network.
input_space = Dict(
{
"img": FloatBox(shape=(6, 6, 3)), # some RGB img
"txt": TextBox() # some text
},
add_batch_rank=True,
add_time_rank=True)
img, txt = ContainerSplitter("img", "txt")(input_space)
# Complex NN assembly via our Keras-style functional API.
# Fold text input into single batch rank.
folded_text = ReShape(fold_time_rank=True)(txt)
# String layer will create batched AND time-ranked (individual words) hash outputs (int64).
string_bucket_out, lengths = StringToHashBucket(
num_hash_buckets=5)(folded_text)
# Batched and time-ranked embedding output (floats) with embed dim=n.
embedding_out = EmbeddingLookup(embed_dim=10,
vocab_size=5)(string_bucket_out)
# Pass embeddings through a text LSTM and use last output (reduce time-rank).
string_lstm_out, _ = LSTMLayer(units=2,
return_sequences=False,
scope="lstm-layer-txt")(
embedding_out,
sequence_length=lengths)
# Unfold to get original time-rank back.
string_lstm_out_unfolded = ReShape(unfold_time_rank=True)(
string_lstm_out, txt)
# Parallel image stream via 1 CNN layer plus dense.
folded_img = ReShape(fold_time_rank=True, scope="img-fold")(img)
cnn_out = Conv2DLayer(filters=1, kernel_size=2, strides=2)(folded_img)
unfolded_cnn_out = ReShape(unfold_time_rank=True,
scope="img-unfold")(cnn_out, img)
unfolded_cnn_out_flattened = ReShape(
flatten=True, scope="img-flat")(unfolded_cnn_out)
dense_out = DenseLayer(units=2,
scope="dense-0")(unfolded_cnn_out_flattened)
# Concat everything.
concat_out = ConcatLayer()(string_lstm_out_unfolded, dense_out)
# LSTM output has batch+time.
main_lstm_out, internal_states = LSTMLayer(
units=2, scope="lstm-layer-main")(concat_out)
dense1_after_lstm_out = DenseLayer(units=3,
scope="dense-1")(main_lstm_out)
dense2_after_lstm_out = DenseLayer(
units=2, scope="dense-2")(dense1_after_lstm_out)
dense3_after_lstm_out = DenseLayer(
units=1, scope="dense-3")(dense2_after_lstm_out)
# A NN with 2 outputs.
neural_net = NeuralNetwork(
outputs=[dense3_after_lstm_out, main_lstm_out, internal_states])
test = ComponentTest(component=neural_net,
input_spaces=dict(inputs=input_space))
# Batch of size=n.
sample_shape = (4, 2)
input_ = input_space.sample(sample_shape)
out = test.test(("call", input_), expected_outputs=None)
# Main output (Dense out after LSTM).
self.assertTrue(out[0].shape == sample_shape +
(1, )) # 1=1 unit in dense layer
self.assertTrue(out[0].dtype == np.float32)
# main-LSTM out.
self.assertTrue(out[1].shape == sample_shape + (2, )) # 2=2 LSTM units
self.assertTrue(out[1].dtype == np.float32)
# main-LSTM internal-states.
self.assertTrue(out[2][0].shape == sample_shape[:1] +
(2, )) # 2=2 LSTM units
self.assertTrue(out[2][0].dtype == np.float32)
self.assertTrue(out[2][1].shape == sample_shape[:1] +
(2, )) # 2=2 LSTM units
self.assertTrue(out[2][1].dtype == np.float32)
test.terminate()
def test_keras_style_complex_multi_stream_nn(self):
# 3 inputs.
input_spaces = [
Dict({
"img": FloatBox(shape=(6, 6, 3)),
"int": IntBox(3)
}, add_batch_rank=True, add_time_rank=True),
FloatBox(shape=(2,), add_batch_rank=True),
Tuple(IntBox(2), TextBox(), add_batch_rank=True, add_time_rank=True)
]
# Same NN as in test above, only using some of the sub-Spaces from the input spaces.
# Tests whether this NN can add automatically the correct splitters.
folded_text = ReShape(fold_time_rank=True)(input_spaces[2][1])
# String layer will create batched AND time-ranked (individual words) hash outputs (int64).
string_bucket_out, lengths = StringToHashBucket(num_hash_buckets=5)(folded_text)
# Batched and time-ranked embedding output (floats) with embed dim=n.
embedding_out = EmbeddingLookup(embed_dim=10, vocab_size=5)(string_bucket_out)
# Pass embeddings through a text LSTM and use last output (reduce time-rank).
string_lstm_out, _ = LSTMLayer(units=2, return_sequences=False, scope="lstm-layer-txt")(
embedding_out, sequence_length=lengths
)
# Unfold to get original time-rank back.
string_lstm_out_unfolded = ReShape(unfold_time_rank=True)(string_lstm_out, input_spaces[2][1])
# Parallel image stream via 1 CNN layer plus dense.
folded_img = ReShape(fold_time_rank=True, scope="img-fold")(input_spaces[0]["img"])
cnn_out = Conv2DLayer(filters=1, kernel_size=2, strides=2)(folded_img)
unfolded_cnn_out = ReShape(unfold_time_rank=True, scope="img-unfold")(cnn_out, input_spaces[0]["img"])
unfolded_cnn_out_flattened = ReShape(flatten=True, scope="img-flat")(unfolded_cnn_out)
dense_out = DenseLayer(units=2, scope="dense-0")(unfolded_cnn_out_flattened)
# Concat everything.
concat_out = ConcatLayer()(string_lstm_out_unfolded, dense_out)
# LSTM output has batch+time.
main_lstm_out, internal_states = LSTMLayer(units=2, scope="lstm-layer-main")(concat_out)
dense1_after_lstm_out = DenseLayer(units=3, scope="dense-1")(main_lstm_out)
dense2_after_lstm_out = DenseLayer(units=2, scope="dense-2")(dense1_after_lstm_out)
dense3_after_lstm_out = DenseLayer(units=1, scope="dense-3")(dense2_after_lstm_out)
# A NN with 3 outputs.
neural_net = NeuralNetwork(inputs=input_spaces, outputs=[dense3_after_lstm_out, main_lstm_out, internal_states])
test = ComponentTest(component=neural_net, input_spaces=dict(inputs=input_spaces))
# Batch of size=n.
sample_shape = (4, 2)
input_ = [input_spaces[0].sample(sample_shape), input_spaces[1].sample(sample_shape[0]),
input_spaces[2].sample(sample_shape)]
out = test.test(("call", tuple(input_)), expected_outputs=None)
# Main output (Dense out after LSTM).
self.assertTrue(out[0].shape == sample_shape + (1,)) # 1=1 unit in dense layer
self.assertTrue(out[0].dtype == np.float32)
# main-LSTM out.
self.assertTrue(out[1].shape == sample_shape + (2,)) # 2=2 LSTM units
self.assertTrue(out[1].dtype == np.float32)
# main-LSTM internal-states.
self.assertTrue(out[2][0].shape == sample_shape[:1] + (2,)) # 2=2 LSTM units
self.assertTrue(out[2][0].dtype == np.float32)
self.assertTrue(out[2][1].shape == sample_shape[:1] + (2,)) # 2=2 LSTM units
self.assertTrue(out[2][1].dtype == np.float32)
test.terminate()