m53297601
/
mindspore
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

!8586 fix GatherD and add Randperm for aicpu

Browse Source 
From: @yanzhenxiang2020
Reviewed-by: @liangchenghui,@wuxuejian
Signed-off-by: @wuxuejian
pull/8586/MERGE
mindspore-ci-bot 5 years ago committed by Gitee
parent 18f58f7db4 082cb4e7fd
commit 2370828043
10 changed files with 371 additions and 7 deletions
Show Stats Download Patch File Download Diff File

2
mindspore/ccsrc/backend/kernel_compiler/aicpu/aicpu_util.h
Unescape View File

@ -47,7 +47,7 @@ constexpr auto kEditDistance = "EditDistance";
constexpr auto kGatherD = "GatherD";
constexpr auto kIdentity = "Identity";
constexpr auto kCustRunApi = "RunCpuKernel";
const std::set<std::string> kCustAiCpuKernelOps{kEditDistance, kGatherD, kIdentity, kMeshgrid};
const std::set<std::string> kCustAiCpuKernelOps{kEditDistance, kIdentity};
struct AicpuParamHead {
uint32_t length; // Total length: include cunstom message

3
mindspore/ops/_op_impl/aicpu/__init__.py
Unescape View File

@ -18,6 +18,8 @@ from .init_data_set_queue import _init_data_set_queue_aicpu
from .embedding_lookup import _embedding_lookup_aicpu
from .padding import _padding_aicpu
from .gather import _gather_aicpu
from .gather_grad import _gather_grad_aicpu
from .scatter import _scatter_aicpu
from .identity import _identity_aicpu
from .edit_distance import _edit_distance_aicpu
from .unique_with_pad import _unique_with_pad_aicpu
@ -30,6 +32,7 @@ from .reshape import _reshape_aicpu
from .flatten import _flatten_aicpu
from .squeeze import _squeeze_aicpu
from .expand_dims import _expand_dims_aicpu
from .randperm import _randperm_aicpu
from .random_choice_with_mask import _random_choice_with_mask_aicpu
from .pack import _pack_aicpu
from .ctcloss import _ctcloss_aicpu

54
mindspore/ops/_op_impl/aicpu/gather_grad.py
Unescape View File

@ -0,0 +1,54 @@
# 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.
# ============================================================================
"""GatherDGrad op"""
from mindspore.ops.op_info_register import op_info_register, AiCPURegOp, DataType
gather_grad_op_info = AiCPURegOp("GatherDGrad") \
.fusion_type("OPAQUE") \
.attr("dim", "int") \
.input(0, "index", "required") \
.input(1, "src", "required") \
.output(0, "output", "required") \
.dtype_format(DataType.I32_Default, DataType.I8_Default, DataType.I8_Default) \
.dtype_format(DataType.I32_Default, DataType.I16_Default, DataType.I16_Default) \
.dtype_format(DataType.I32_Default, DataType.I32_Default, DataType.I32_Default) \
.dtype_format(DataType.I32_Default, DataType.I64_Default, DataType.I64_Default) \
.dtype_format(DataType.I32_Default, DataType.U8_Default, DataType.U8_Default) \
.dtype_format(DataType.I32_Default, DataType.U16_Default, DataType.U16_Default) \
.dtype_format(DataType.I32_Default, DataType.U32_Default, DataType.U32_Default) \
.dtype_format(DataType.I32_Default, DataType.U64_Default, DataType.U64_Default) \
.dtype_format(DataType.I32_Default, DataType.F16_Default, DataType.F16_Default) \
.dtype_format(DataType.I32_Default, DataType.F32_Default, DataType.F32_Default) \
.dtype_format(DataType.I32_Default, DataType.F64_Default, DataType.F64_Default) \
.dtype_format(DataType.I32_Default, DataType.BOOL_Default, DataType.BOOL_Default) \
.dtype_format(DataType.I64_Default, DataType.I8_Default, DataType.I8_Default) \
.dtype_format(DataType.I64_Default, DataType.I16_Default, DataType.I16_Default) \
.dtype_format(DataType.I64_Default, DataType.I32_Default, DataType.I32_Default) \
.dtype_format(DataType.I64_Default, DataType.I64_Default, DataType.I64_Default) \
.dtype_format(DataType.I64_Default, DataType.U8_Default, DataType.U8_Default) \
.dtype_format(DataType.I64_Default, DataType.U16_Default, DataType.U16_Default) \
.dtype_format(DataType.I64_Default, DataType.U32_Default, DataType.U32_Default) \
.dtype_format(DataType.I64_Default, DataType.U64_Default, DataType.U64_Default) \
.dtype_format(DataType.I64_Default, DataType.F16_Default, DataType.F16_Default) \
.dtype_format(DataType.I64_Default, DataType.F32_Default, DataType.F32_Default) \
.dtype_format(DataType.I64_Default, DataType.F64_Default, DataType.F64_Default) \
.dtype_format(DataType.I64_Default, DataType.BOOL_Default, DataType.BOOL_Default) \
.get_op_info()
@op_info_register(gather_grad_op_info)
def _gather_grad_aicpu():
"""GatherDGrad AiCPU register"""
return

36
mindspore/ops/_op_impl/aicpu/randperm.py
Unescape View File

@ -0,0 +1,36 @@
# 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.
# ============================================================================
"""Randperm op"""
from mindspore.ops.op_info_register import op_info_register, AiCPURegOp, DataType
randperm_op_info = AiCPURegOp("Randperm") \
.fusion_type("OPAQUE") \
.output(0, "y", "required") \
.attr("n", "int") \
.dtype_format(DataType.I8_Default) \
.dtype_format(DataType.I16_Default) \
.dtype_format(DataType.I32_Default) \
.dtype_format(DataType.I64_Default) \
.dtype_format(DataType.U8_Default) \
.dtype_format(DataType.U16_Default) \
.dtype_format(DataType.U32_Default) \
.dtype_format(DataType.U64_Default) \
.get_op_info()
@op_info_register(randperm_op_info)
def _randperm_aicpu():
"""Randperm AiCPU register"""
return

79
mindspore/ops/_op_impl/aicpu/scatter.py
Unescape View File

@ -0,0 +1,79 @@
# 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.
# ============================================================================
"""Scatter op"""
from mindspore.ops.op_info_register import op_info_register, AiCPURegOp, DataType
scatter_op_info = AiCPURegOp("Scatter") \
.fusion_type("OPAQUE") \
.input(0, "target", "required") \
.input(1, "dim", "required") \
.input(2, "index", "required") \
.input(3, "src", "required") \
.output(0, "output", "required") \
.dtype_format(DataType.I8_Default, DataType.I32_Default, \
DataType.I32_Default, DataType.I8_Default, DataType.I8_Default) \
.dtype_format(DataType.I16_Default, DataType.I32_Default, \
DataType.I32_Default, DataType.I16_Default, DataType.I16_Default) \
.dtype_format(DataType.I32_Default, DataType.I32_Default, \
DataType.I32_Default, DataType.I32_Default, DataType.I32_Default) \
.dtype_format(DataType.I64_Default, DataType.I32_Default, \
DataType.I32_Default, DataType.I64_Default, DataType.I64_Default) \
.dtype_format(DataType.U8_Default, DataType.I32_Default, \
DataType.I32_Default, DataType.U8_Default, DataType.U8_Default) \
.dtype_format(DataType.U16_Default, DataType.I32_Default, \
DataType.I32_Default, DataType.U16_Default, DataType.U16_Default) \
.dtype_format(DataType.U32_Default, DataType.I32_Default, \
DataType.I32_Default, DataType.U32_Default, DataType.U32_Default) \
.dtype_format(DataType.U64_Default, DataType.I32_Default, \
DataType.I32_Default, DataType.U64_Default, DataType.U64_Default) \
.dtype_format(DataType.F16_Default, DataType.I32_Default, \
DataType.I32_Default, DataType.F16_Default, DataType.F16_Default) \
.dtype_format(DataType.F32_Default, DataType.I32_Default, \
DataType.I32_Default, DataType.F32_Default, DataType.F32_Default) \
.dtype_format(DataType.F64_Default, DataType.I32_Default, \
DataType.I32_Default, DataType.F64_Default, DataType.F64_Default) \
.dtype_format(DataType.BOOL_Default, DataType.I32_Default, \
DataType.I32_Default, DataType.BOOL_Default, DataType.BOOL_Default) \
.dtype_format(DataType.I8_Default, DataType.I32_Default, \
DataType.I64_Default, DataType.I8_Default, DataType.I8_Default) \
.dtype_format(DataType.I16_Default, DataType.I32_Default, \
DataType.I64_Default, DataType.I16_Default, DataType.I16_Default) \
.dtype_format(DataType.I32_Default, DataType.I32_Default, \
DataType.I64_Default, DataType.I32_Default, DataType.I32_Default) \
.dtype_format(DataType.I64_Default, DataType.I32_Default, \
DataType.I64_Default, DataType.I64_Default, DataType.I64_Default) \
.dtype_format(DataType.U8_Default, DataType.I32_Default, \
DataType.I64_Default, DataType.U8_Default, DataType.U8_Default) \
.dtype_format(DataType.U16_Default, DataType.I32_Default, \
DataType.I64_Default, DataType.U16_Default, DataType.U16_Default) \
.dtype_format(DataType.U32_Default, DataType.I32_Default, \
DataType.I64_Default, DataType.U32_Default, DataType.U32_Default) \
.dtype_format(DataType.U64_Default, DataType.I32_Default, \
DataType.I64_Default, DataType.U64_Default, DataType.U64_Default) \
.dtype_format(DataType.F16_Default, DataType.I32_Default, \
DataType.I64_Default, DataType.F16_Default, DataType.F16_Default) \
.dtype_format(DataType.F32_Default, DataType.I32_Default, \
DataType.I64_Default, DataType.F32_Default, DataType.F32_Default) \
.dtype_format(DataType.F64_Default, DataType.I32_Default, \
DataType.I64_Default, DataType.F64_Default, DataType.F64_Default) \
.dtype_format(DataType.BOOL_Default, DataType.I32_Default, \
DataType.I64_Default, DataType.BOOL_Default, DataType.BOOL_Default) \
.get_op_info()
@op_info_register(scatter_op_info)
def _scatter_aicpu():
"""Scatter AiCPU register"""
return

3
mindspore/ops/operations/__init__.py
Unescape View File

@ -56,7 +56,7 @@ from .math_ops import (Abs, ACos, Asin, Asinh, AddN, AccumulateNV2, AssignAdd, A
Sin, Sqrt, Rsqrt, BesselI0e, BesselI1e, TruncateDiv, TruncateMod,
Square, Sub, TensorAdd, Sign, Round, SquareSumAll, Atan, Atanh, Cosh, Sinh, Eps, Tan)
from .random_ops import (RandomChoiceWithMask, StandardNormal, Gamma, Poisson, UniformInt, UniformReal,
from .random_ops import (Randperm, RandomChoiceWithMask, StandardNormal, Gamma, Poisson, UniformInt, UniformReal,
RandomCategorical, StandardLaplace, Multinomial, UniformCandidateSampler)
from .nn_ops import (LSTM, SGD, Adam, FusedSparseAdam, FusedSparseLazyAdam, AdamNoUpdateParam, ApplyMomentum, BatchNorm,
BiasAdd, Conv2D,
@ -198,6 +198,7 @@ __all__ = [
'HSwish',
'HSigmoid',
'Tanh',
'Randperm',
'RandomChoiceWithMask',
'StandardNormal',
'Multinomial',

6
mindspore/ops/operations/array_ops.py
Unescape View File

@ -4601,12 +4601,8 @@ class GatherD(PrimitiveWithInfer):
idx_rank = len(idx_shp)
validator.check("x_rank, idx_rank", x_rank, "expected", idx_rank, Rel.EQ, self.name)
dim_v = dim['value']
validator.check("dim value", dim_v, "expected", 0, Rel.GE, self.name)
validator.check("dim value", dim_v, "expected", -x_rank, Rel.GE, self.name)
validator.check("dim value", dim_v, "expected", x_rank, Rel.LT, self.name)
for i in range(x_rank):
if i == dim_v:
continue
validator.check("x_shp[{0}], idx_shp[{0}]".format(i), x_shp[i], "expected", idx_shp[i], Rel.EQ, self.name)
out = {'shape': index['shape'],
'dtype': x['dtype'],

40
mindspore/ops/operations/random_ops.py
Unescape View File

@ -345,6 +345,46 @@ class UniformReal(PrimitiveWithInfer):
return out
class Randperm(PrimitiveWithInfer):
"""
Generates random samples from 0 to n-1.
Args:
n (int): Number of items expected to get and the number must be greater than 0. Default: 1.
dtype (mindspore.dtype): The type of output. Default: mindspore.int32.
Outputs:
- **output** (Tensor) - The output Tensor with shape :math:`(n,)` and type: dtype.
Supported Platforms:
``Ascend``
Examples:
>>> randperm = ops.Randperm(20)
>>> output = randperm()
>>> print(output)
[15 6 11 19 14 16 9 5 13 18 4 10 8 0 17 2 14 1 12 3 7]
"""
@prim_attr_register
def __init__(self, n=1, dtype=mstype.int32):
"""Initialize Randperm"""
Validator.check_value_type("n", n, [int], self.name)
self.dtype = dtype
self.n = n
self.init_prim_io_names(inputs=[], outputs=['output'])
def infer_shape(self):
Validator.check_int(self.n, 1, Rel.GE, "1", self.name)
return [self.n]
def infer_dtype(self):
valid_values = (mstype.int8, mstype.int16, mstype.int32, mstype.int64,
mstype.uint8, mstype.uint16, mstype.uint32, mstype.uint64)
Validator.check_type_name("dtype", self.dtype, valid_values, self.name)
return self.dtype
class RandomChoiceWithMask(PrimitiveWithInfer):
"""
Generates a random sample as index tensor with a mask tensor from a given tensor.

99
tests/st/ops/ascend/test_aicpu_ops/test_gather_d.py
Unescape View File

@ -0,0 +1,99 @@
# 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.
# ============================================================================
import numpy as np
import mindspore
import mindspore.nn as nn
import mindspore.context as context
from mindspore import Tensor
from mindspore.ops import operations as P
from mindspore.ops.operations import _grad_ops as G
context.set_context(mode=context.GRAPH_MODE, device_target="Ascend")
class Net(nn.Cell):
def __init__(self, dim=0):
super(Net, self).__init__()
self.op = P.GatherD()
self.dim = dim
def construct(self, x, index):
return self.op(x, self.dim, index)
class NetGrad(nn.Cell):
def __init__(self, dim=0, shape=None):
super(NetGrad, self).__init__()
self.op = G.GatherDGrad(dim, shape)
def construct(self, index, x):
return self.op(index, x)
def test_net():
x = Tensor(np.array([[772, 231, 508, 545, 615, 249],
[923, 210, 480, 696, 482, 761],
[465, 904, 521, 824, 607, 669],
[156, 539, 56, 159, 916, 566],
[122, 676, 714, 261, 19, 936]]), mindspore.int32)
index = Tensor(np.array([[0, 0, 0, 1, 1],
[0, 0, 0, 1, 4],
[0, 0, 0, 1, -1],
[1, 1, 1, 0, 0]]), mindspore.int32)
dim = 0
net = Net(dim)
out = net(x, index)
print(out.asnumpy())
expect_out = np.array([[772, 231, 508, 696, 482],
[772, 231, 508, 696, 19],
[772, 231, 508, 696, 19],
[923, 210, 480, 545, 615]])
assert np.array_equal(out.asnumpy(), expect_out)
def test_net_bool():
x = Tensor(np.array([[0, 1, 0, 0, 1, 0],
[0, 1, 0, 0, 1, 0],
[0, 0, 1, 1, 0, 1],
[1, 0, 1, 1, 0, 0],
[1, 1, 1, 1, 0, 0]]), mindspore.bool_)
index = Tensor(np.array([[0, 0, 0, 1, 1],
[0, 0, 0, 1, 4],
[0, 0, 0, 1, -1],
[1, 1, 1, 0, 0]]), mindspore.int32)
dim = 0
net = Net(dim)
out = net(x, index)
print(out.asnumpy())
expect_out = np.array([[0, 1, 0, 0, 1],
[0, 1, 0, 0, 0],
[0, 1, 0, 0, 0],
[0, 1, 0, 0, 1]]).astype(np.bool)
assert np.array_equal(out.asnumpy(), expect_out)
def test_net_grad():
index = Tensor(np.array([[0, 1, 2, 0, 0],
[2, 0, 0, 1, -1]]), mindspore.int32)
x = Tensor(np.array([[772, 231, 508, 615, 249],
[122, 676, 714, 261, 936]]), mindspore.int32)
net = NetGrad(dim=0, shape=(3, 5))
out = net(index, x)
print(out.asnumpy())
expect_out = np.array([[772, 676, 714, 615, 249],
[0, 231, 0, 261, 0],
[122, 0, 508, 0, 936]])
assert np.array_equal(out.asnumpy(), expect_out)

56
tests/st/ops/ascend/test_aicpu_ops/test_randperm.py
Unescape View File

@ -0,0 +1,56 @@
# 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.
# ============================================================================
import mindspore
import mindspore.nn as nn
import mindspore.context as context
from mindspore.ops import operations as P
context.set_context(mode=context.GRAPH_MODE, device_target="Ascend")
class Net(nn.Cell):
def __init__(self, n=1, dtype=mindspore.int32):
super(Net, self).__init__()
self.randperm = P.Randperm(n, dtype)
def construct(self):
return self.randperm()
def test_net():
net = Net()
output = net()
print(output)
print(output.shape)
print(output.dtype)
assert output.shape == (1,)
assert output.dtype == mindspore.int32
assert output.asnumpy()[0] == 0
def test_net_n20():
net = Net(20, mindspore.uint64)
output = net()
print(output)
assert output.shape == (20,)
assert output.dtype == mindspore.uint64
sample_set = set()
for i in output.asnumpy():
assert i not in sample_set
assert 0 <= i < 20
sample_set.add(i)
Write Preview
Loading…
Cancel
Save