p60975123
/
mindspore
forked from mindspore-Ecosystem/mindspore
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

!14013 Enable zero dimension using attribute not parameter 

From: @liangzhibo
Reviewed-by: @zh_qh,@ginfung
Signed-off-by: @zh_qh
This commit is contained in:
mindspore-ci-bot 2021-03-25 17:49:35 +08:00 committed by Gitee
parent c0a6117798 af6d812436
commit c1a802627b
2 changed files with 16 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

15
mindspore/common/tensor.py
View File

@ -74,7 +74,7 @@ class Tensor(Tensor_):
>>> assert t3.dtype == ms.float32
"""
def __init__(self, input_data=None, dtype=None, shape=None, init=None, check_zero_dims=True):
def __init__(self, input_data=None, dtype=None, shape=None, init=None):
self.init_finished = False
# If input data is numpy number, convert it to np array
if isinstance(input_data, np_types):
@ -92,13 +92,12 @@ class Tensor(Tensor_):
if isinstance(shape, numbers.Number):
shape = (shape,)
if check_zero_dims:
if input_data is not None and isinstance(input_data, (tuple, list, np.ndarray)) \
and np.array(input_data).ndim > 1 and np.array(input_data).size == 0:
raise ValueError("input_data can not contain zero dimension.")
if shape is not None:
if 0 in shape:
raise ValueError("Shape can not contain zero value.")
if input_data is not None and isinstance(input_data, (tuple, list, np.ndarray)) \
and np.array(input_data).ndim > 1 and np.array(input_data).size == 0:
raise ValueError("input_data can not contain zero dimension.")
if shape is not None and not (hasattr(init, "__enable_zero_dim__") and init.__enable_zero_dim__):
if 0 in shape:
raise ValueError("Shape can not contain zero value.")
# If input_data is tuple/list/numpy.ndarray, it's support in check_type method.
if init is None:

11
mindspore/ops/operations/array_ops.py
View File

@ -26,6 +26,7 @@ import numbers
import numpy as np
from mindspore import log as logger
from mindspore.common.initializer import Zero
from .._utils import get_concat_offset
from ..operations.math_ops import _infer_shape_reduce
from ..primitive import Primitive, PrimitiveWithInfer, PrimitiveWithCheck, prim_attr_register, _run_op
@ -38,6 +39,7 @@ from ...common.parameter import Parameter
from ...common.tensor import Tensor
class _ScatterOp(PrimitiveWithInfer):
"""
Defines Scatter operators
@ -3015,8 +3017,13 @@ class StridedSlice(PrimitiveWithInfer):
ret_shape = self._compute_slicing_shape(x['shape'], begin_v, end_v, strides_v)
value = None if all(ret_shape) else Tensor(np.array([]).reshape(ret_shape), x['dtype'].element_type(),
check_zero_dims=False)
if all(ret_shape):
value = None
else:
init_func = Zero()
init_func.__enable_zero_dim__ = True
value = Tensor(dtype=x['dtype'].element_type(), shape=ret_shape, init=init_func)
if "max_value" in x and "min_value" in x:
validator.check_value_type("min_value", x["min_value"], [tuple, list], self.name)
validator.check_value_type("max_value", x["max_value"], [tuple, list], self.name)