experiments/STT-tensorflow
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Fix bug in RaggedTensor broadcasting that impacted broadcasting inner dense dimensions.

Browse Source 
PiperOrigin-RevId: 330564050
Change-Id: I2226c49996cd386260c999906ed6fdf5cf8da417
This commit is contained in:
Edward Loper 2020-09-08 12:37:41 -07:00 committed by TensorFlower Gardener
parent f67204eea0
commit 2db7fda9dd
2 changed files with 40 additions and 34 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
tensorflow/python/ops/ragged/ragged_tensor_shape.py
View File

@ -582,10 +582,12 @@ def _broadcast_to_ragged_shape(rt_input, dst_shape, broadcast_inner_dimensions):
rt_input = ragged_array_ops.tile(rt_input, multiples)
if broadcast_inner_dimensions:
new_shape = array_ops.broadcast_dynamic_shape(
array_ops.shape(
rt_input.flat_values, out_type=dst_shape.dim_size_dtype),
array_ops.concat([[1], dst_shape.inner_dim_sizes], axis=0))
rt_input = rt_input.with_flat_values(
array_ops.reshape(
rt_input.flat_values,
array_ops.concat([[-1], dst_shape.inner_dim_sizes], axis=0)))
array_ops.broadcast_to(rt_input.flat_values, new_shape))
# Do broadcasting for dimensions that become ragged. We must do these from
# outermost to innermost.

66
tensorflow/python/ops/ragged/ragged_tensor_shape_test.py
View File

@ -378,37 +378,41 @@ class RaggedTensorShapeTest(test_util.TensorFlowTestCase,
r'partitioned_dim_sizes=\(<[^>]+>, <[^>]+>\), '
r'inner_dim_sizes=<[^>]+>\)')
@parameterized.parameters(
[
dict(
x=[[10], [20], [30]], # shape=[3, 1]
dim_sizes=[3, 2],
expected=[[10, 10], [20, 20], [30, 30]]),
dict(
x=[[10], [20], [30]], # shape=[3, 1]
dim_sizes=[3, [3, 0, 2]],
expected=ragged_factory_ops.constant_value(
[[10, 10, 10], [], [30, 30]], dtype=np.int32)),
dict(
x=[[[1, 2, 3]], [[4, 5, 6]]], # shape = [2, 1, 3]
dim_sizes=[2, [2, 3], 3],
expected=ragged_factory_ops.constant_value(
[[[1, 2, 3], [1, 2, 3]], [[4, 5, 6], [4, 5, 6], [4, 5, 6]]],
dtype=np.int32,
ragged_rank=1)),
dict(
x=[[[1]], [[2]]], # shape = [2, 1, 1]
dim_sizes=[2, [2, 3], [0, 2, 1, 2, 0]],
expected=ragged_factory_ops.constant_value(
[[[], [1, 1]], [[2], [2, 2], []]],
dtype=np.int32,
ragged_rank=2)),
dict(
x=10,
dim_sizes=[3, [3, 0, 2]],
expected=ragged_factory_ops.constant_value([[10, 10, 10], [],
[10, 10]])),
])
@parameterized.parameters([
dict(
x=[[10], [20], [30]], # shape=[3, 1]
dim_sizes=[3, 2],
expected=[[10, 10], [20, 20], [30, 30]]),
dict(
x=[[10], [20], [30]], # shape=[3, 1]
dim_sizes=[3, [3, 0, 2]],
expected=ragged_factory_ops.constant_value(
[[10, 10, 10], [], [30, 30]], dtype=np.int32)),
dict(
x=[[[1, 2, 3]], [[4, 5, 6]]], # shape = [2, 1, 3]
dim_sizes=[2, [2, 3], 3],
expected=ragged_factory_ops.constant_value(
[[[1, 2, 3], [1, 2, 3]], [[4, 5, 6], [4, 5, 6], [4, 5, 6]]],
dtype=np.int32,
ragged_rank=1)),
dict(
x=[[[1]], [[2]]], # shape = [2, 1, 1]
dim_sizes=[2, [2, 3], [0, 2, 1, 2, 0]],
expected=ragged_factory_ops.constant_value(
[[[], [1, 1]], [[2], [2, 2], []]], dtype=np.int32,
ragged_rank=2)),
dict(
x=10,
dim_sizes=[3, [3, 0, 2]],
expected=ragged_factory_ops.constant_value([[10, 10, 10], [],
[10, 10]])),
dict(
x=ragged_factory_ops.constant_value([[[1], [2]], [[3]]],
ragged_rank=1),
dim_sizes=[2, [2, 1], 2],
expected=ragged_factory_ops.constant_value(
[[[1, 1], [2, 2]], [[3, 3]]], ragged_rank=1)),
])
def testRaggedBroadcastTo(self, x, dim_sizes, expected):
shape = RaggedTensorDynamicShape.from_dim_sizes(dim_sizes)
result = ragged_tensor_shape.broadcast_to(x, shape)