!4789 Add EditDistance op for GE.

Merge pull request !4789 from liuxiao93/Add-EditDistance-op-for-GE

mindspore/ccsrc/transform/graph_ir/op_adapter_map.h

@@ -190,6 +190,7 @@ constexpr const char kNameSquareSumAll[] = "SquareSumAll";
 constexpr const char kNameAscendQuant[] = "Quant";
 constexpr const char kNameAscendDequant[] = "Dequant";
 constexpr const char kNameReverseSequence[] = "ReverseSequence";
+constexpr const char kNameEditDistance[] = "EditDistance";
 constexpr const char kNameCase[] = "Case";
 
 class OpAdapterMap {

mindspore/ccsrc/transform/graph_ir/op_declare/array_ops_declare.cc

@@ -87,4 +87,12 @@ ATTR_MAP(ReverseSequence) = {{"seq_dim", ATTR_DESC(seq_dim, AnyTraits<int>())},
                              {"batch_dim", ATTR_DESC(batch_dim, AnyTraits<int>())}};
 OUTPUT_MAP(ReverseSequence) = {{0, OUTPUT_DESC(y)}};
 REG_ADPT_DESC(ReverseSequence, kNameReverseSequence, ADPT_DESC(ReverseSequence))
+
+// EditDistance
+INPUT_MAP(EditDistance) = {{1, INPUT_DESC(hypothesis_indices)}, {2, INPUT_DESC(hypothesis_values)},
+                           {3, INPUT_DESC(hypothesis_shape)},   {4, INPUT_DESC(truth_indices)},
+                           {5, INPUT_DESC(truth_values)},       {6, INPUT_DESC(truth_shape)}};
+ATTR_MAP(EditDistance) = {{"normalize", ATTR_DESC(normalize, AnyTraits<bool>())}};
+OUTPUT_MAP(EditDistance) = {{0, OUTPUT_DESC(output)}};
+REG_ADPT_DESC(EditDistance, kNameEditDistance, ADPT_DESC(EditDistance))
 }  // namespace mindspore::transform

mindspore/ccsrc/transform/graph_ir/op_declare/array_ops_declare.h

@@ -54,5 +54,8 @@ DECLARE_OP_ADAPTER(Data)
 
 DECLARE_OP_ADAPTER(ReverseSequence)
 DECLARE_OP_USE_OUTPUT(ReverseSequence)
+
+DECLARE_OP_ADAPTER(EditDistance)
+DECLARE_OP_USE_OUTPUT(EditDistance)
 }  // namespace mindspore::transform
 #endif  // MINDSPORE_CCSRC_TRANSFORM_GRAPH_IR_OP_DECLARE_ARRAY_OPS_DECLARE_H_

mindspore/ops/_op_impl/tbe/__init__.py

@@ -23,7 +23,6 @@ from .acos_grad import _acos_grad_tbe
 from .acosh import _acosh_tbe
 from .acosh_grad import _acosh_grad_tbe
 from .adam_apply_one_with_decay import _adam_apply_one_with_decay_tbe
-from .add import _add_tbe
 from .apply_centered_rms_prop import _apply_centered_rms_prop_tbe
 from .add_n import _add_n_tbe
 from .accumulate_n_v2 import _accumulate_n_v2_tbe

mindspore/ops/_op_impl/tbe/add.py

@@ -1,37 +0,0 @@
-# Copyright 2020 Huawei Technologies Co., Ltd
-#
-# Licensed under the Apache License, Version 2.0 (the "License");
-# you may not use this file except in compliance with the License.
-# You may obtain a copy of the License at
-#
-# http://www.apache.org/licenses/LICENSE-2.0
-#
-# Unless required by applicable law or agreed to in writing, software
-# distributed under the License is distributed on an "AS IS" BASIS,
-# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-# See the License for the specific language governing permissions and
-# limitations under the License.
-# ============================================================================
-
-"""Add op"""
-from mindspore.ops.op_info_register import op_info_register, TBERegOp, DataType
-
-add_op_info = TBERegOp("Add") \
-    .fusion_type("ELEMWISE") \
-    .async_flag(False) \
-    .binfile_name("add.so") \
-    .compute_cost(10) \
-    .kernel_name("add") \
-    .partial_flag(True) \
-    .op_pattern("dynamicFormat") \
-    .input(0, "x1", False, "required", "all") \
-    .input(1, "x2", False, "required", "all") \
-    .output(0, "y", False, "required", "all") \
-    .dtype_format(DataType.None_None, DataType.None_None, DataType.None_None) \
-    .get_op_info()
-
-
-@op_info_register(add_op_info)
-def _add_tbe():
-    """Add TBE register"""
-    return

mindspore/ops/operations/__init__.py

@@ -29,7 +29,7 @@ from .array_ops import (Argmax, Argmin, Cast, Concat, Pack, Unpack,
                         ScatterUpdate, ScalarToArray, ScalarToTensor, ScatterNd, ScatterNdUpdate, Select,
                         Shape, Size, Slice, Split, TransShape, ParallelConcat, Padding,
                         ScatterNdAdd, ScatterNdSub, ScatterNonAliasingAdd, ReverseV2, Rint,
-                        Squeeze, StridedSlice, Tile, TensorScatterUpdate,
+                        Squeeze, StridedSlice, Tile, TensorScatterUpdate, EditDistance,
                         Transpose, TruncatedNormal, TupleToArray, UnsortedSegmentMin, UnsortedSegmentProd,
                         UnsortedSegmentSum, SpaceToDepth, DepthToSpace, SpaceToBatch, BatchToSpace,
                         SpaceToBatchND, BatchToSpaceND, BroadcastTo, InplaceUpdate, ReverseSequence, EmbeddingLookup,
@@ -92,6 +92,7 @@ from .sparse_ops import SparseToDense
 
 __all__ = [
     'ReverseSequence',
+    'EditDistance',
     'CropAndResize',
     'TensorAdd',
     'Argmax',

mindspore/ops/operations/array_ops.py

@@ -3470,6 +3470,93 @@ class ReverseSequence(PrimitiveWithInfer):
         return x
 
 
+class EditDistance(PrimitiveWithInfer):
+    """
+    Computes the Levebshtein Edit Distance. It is used to measure the similarity of two sequences.
+
+    Args:
+        normalize (bool): If True, edit distances are normalized by length of truth. Default: True.
+
+    Inputs:
+        - **hypothesis_indices** (Tensor) - The indices of the hypothesis list SparseTensor. With int64 data type.
+          The shape of tensor is :math:`(N, R)`.
+        - **hypothesis_values** (Tensor) - The values of the hypothesis list SparseTensor.
+          Must be 1-D vector with length of N.
+        - **hypothesis_shape** (Tensor) - The values of the hypothesis list SparseTensor.
+          Must be R-length vector with int64 data type. Only constant value is allowed.
+        - **truth_indices** (Tensor) - The indices of the truth list SparseTensor. With int64 data type.
+          The shape of tensor is :math:`(M, R)`.
+        - **truth_values** (Tensor) - The values of the truth list SparseTensor. Must be 1-D vector with length of M.
+        - **truth_shape** (Tensor) - The values of the truth list SparseTensor.
+          Must be R-length vector with int64 data type. Only constant value is allowed.
+
+    Outputs:
+        Tensor, a dense tensor with rank `R-1` and float32 data type.
+
+    Examples:
+        >>> class EditDistance(nn.Cell):
+        >>>     def __init__(self, hypothesis_shape, truth_shape, normalize=True):
+        >>>         super(EditDistance, self).__init__()
+        >>>         self.edit_distance = P.EditDistance(normalize)
+        >>>         self.hypothesis_shape = hypothesis_shape
+        >>>         self.truth_shape = truth_shape
+        >>>
+        >>>     def construct(self, hypothesis_indices, hypothesis_values, truth_indices, truth_values):
+        >>>         return self.edit_distance(hypothesis_indices, hypothesis_values, self.hypothesis_shape,
+        >>>                                   truth_indices, truth_values, self.truth_shape)
+        >>>
+        >>> hypothesis_indices = Tensor(np.array([[0, 0, 0], [1, 0, 1], [1, 1, 1]]).astype(np.int64))
+        >>> hypothesis_values = Tensor(np.array([1, 2, 3]).astype(np.float32))
+        >>> hypothesis_shape = Tensor(np.array([1, 1, 2]).astype(np.int64))
+        >>> truth_indices = Tensor(np.array([[0, 1, 0], [0, 0, 1], [1, 1, 0], [1, 0, 1]]).astype(np.int64))
+        >>> truth_values = Tensor(np.array([1, 3, 2, 1]).astype(np.float32))
+        >>> truth_shape = Tensor(np.array([2, 2, 2]).astype(np.int64))
+        >>> edit_distance = EditDistance(hypothesis_shape, truth_shape)
+        >>> out = edit_distance(hypothesis_indices, hypothesis_values, truth_indices, truth_values)
+        >>> [[1.0, 1.0], [1.0, 1.0]]
+    """
+
+    @prim_attr_register
+    def __init__(self, normalize=True):
+        """init EditDistance"""
+        self.normalize = validator.check_value_type("normalize", normalize, [bool], self.name)
+
+    def __infer__(self, h_indices, h_values, h_shape, truth_indices, truth_values, truth_shape):
+        validator.check_const_input('hypothesis_shape', h_shape['value'], self.name)
+        validator.check_const_input('truth_shape', truth_shape['value'], self.name)
+        args_int = {"hypothesis_indices": h_indices['dtype'], "hypothesis_shape": h_shape['dtype'],
+                    "truth_indices": truth_indices['dtype'], "truth_shape": truth_shape['dtype']}
+        validator.check_tensor_type_same(args_int, [mstype.int64], self.name)
+        args = {"hypothesis_values": h_values['dtype'], "truth_values": truth_values['dtype']}
+        validator.check_tensor_type_same(args, mstype.number_type, self.name)
+
+        hypothesis_indices_shp, truth_indices_shp = h_indices['shape'], truth_indices['shape']
+        validator.check("hypothesis_indices rank", len(hypothesis_indices_shp), "expected", 2, Rel.EQ, self.name)
+        validator.check("truth_indices rank", len(truth_indices_shp), "expected", 2, Rel.EQ, self.name)
+        validator.check("hypothesis_values rank", len(h_values['shape']), "expected", 1, Rel.EQ, self.name)
+        validator.check("hypothesis_shape rank", len(h_shape['shape']), "expected", 1, Rel.EQ, self.name)
+        validator.check("truth_values rank", len(truth_values['shape']), "expected", 1, Rel.EQ, self.name)
+        validator.check("truth_shape rank", len(truth_shape['shape']), "expected", 1, Rel.EQ, self.name)
+        validator.check("hypothesis_values shape", h_values['shape'][0],
+                        "hypothesis_indices shape[0]", hypothesis_indices_shp[0], Rel.EQ, self.name)
+        validator.check("hypothesis_shape", h_shape['shape'][0],
+                        "hypothesis_indices shape[1]", hypothesis_indices_shp[1], Rel.EQ, self.name)
+        validator.check("truth_values shape", truth_values['shape'][0],
+                        "truth_indices shape[0]", truth_indices_shp[0], Rel.EQ, self.name)
+        validator.check("hypothesis_shape", h_shape['shape'][0],
+                        "truth_shape", truth_shape['shape'][0], Rel.EQ, self.name)
+        hypothesis_shape_v = h_shape['value'].asnumpy()
+        truth_shape_v = truth_shape['value'].asnumpy()
+        out_shape_rank = len(hypothesis_shape_v) - 1
+        out_shape = []
+        for i in range(out_shape_rank):
+            out_shape.append(max(hypothesis_shape_v[i], truth_shape_v[i]))
+
+        return {'shape': tuple(out_shape),
+                'dtype': mstype.tensor_type(mstype.float32),
+                'value': None}
+
+
 class TransShape(PrimitiveWithInfer):
     """
     Transform the shape of input tensor to target shape.

tests/ut/python/ops/test_ops.py

@@ -684,6 +684,18 @@ class ParallelConcatNet(nn.Cell):
         return self.parallel_concat((x1, x2))
 
 
+class EditDistance(nn.Cell):
+    def __init__(self, hypothesis_shape, truth_shape, normalize=True):
+        super(EditDistance, self).__init__()
+        self.edit_distance = P.EditDistance(normalize)
+        self.hypothesis_shape = hypothesis_shape
+        self.truth_shape =truth_shape
+
+    def construct(self, hypothesis_indices, hypothesis_values, truth_indices, truth_values):
+        return self.edit_distance(hypothesis_indices, hypothesis_values, self.hypothesis_shape,
+                                  truth_indices, truth_values, self.truth_shape)
+
+
 test_case_math_ops = [
     ('BitwiseAnd', {
         'block': P.BitwiseAnd(),
@@ -1978,6 +1990,15 @@ test_case_array_ops = [
         'desc_inputs': [Tensor(np.array([[1, 2, 3], [4, 5, 6], [7, 8, 9]]).astype(np.float32)),
                         Tensor(np.array([1, 2, 3]).astype(np.int32))],
         'desc_bprop': [[3, 3]]}),
+    ('EditDistance', {
+        'block': EditDistance(Tensor(np.array([1, 1, 2]).astype(np.int64)),
+                              Tensor(np.array([2, 2, 2]).astype(np.int64))),
+        'desc_inputs': [Tensor(np.array([[0, 0, 0], [1, 0, 1], [1, 1, 1]]).astype(np.int64)),
+                        Tensor(np.array([1, 2, 3]).astype(np.float32)),
+                        Tensor(np.array([[0, 1, 0], [0, 0, 1], [1, 1, 0], [1, 0, 1]]).astype(np.int64)),
+                        Tensor(np.array([1, 3, 2, 1]).astype(np.float32))],
+        'skip': ['backward'],
+    }),
     ('LinSpace', {
         'block': inner.LinSpace(),
         'desc_inputs': [Tensor([5, 5.5], mstype.float32),