未验证 提交 546b1c1d 编写于 作者: H Huihuang Zheng 提交者: GitHub

Fix paddle.jit.dy2static.data_layer_not_check 2.0 API Doc (#28461)

Remove "fluid", "variable" in 2.0 API doc
上级 8f664a5c
...@@ -29,19 +29,19 @@ __all__ = [ ...@@ -29,19 +29,19 @@ __all__ = [
def data_layer_not_check(name, shape, dtype='float32', lod_level=0): def data_layer_not_check(name, shape, dtype='float32', lod_level=0):
""" """
This function creates a variable on the global block. Unlike This function creates a Tensor on the global block. The created Tensor
`paddle.fluid.data` , the created variable doesn't check the dtype and the doesn't check the dtype and the shape of feed data because dygraph input
shape of feed data because dygraph input data can be variable-length. data can be various-length. This API is used in translating dygraph into
This API is used in translating dygraph into static graph. static graph.
Note: Note:
The default :code:`stop_gradient` attribute of the Variable created by The default :code:`stop_gradient` attribute of the Tensor created by
this API is true, which means the gradient won't be passed backward this API is true, which means the gradient won't be passed backward
through the data Varaible. Set :code:`var.stop_gradient = False` If through the data Tensor. Set :code:`var.stop_gradient = False` If
user would like to pass backward gradient. user would like to pass backward gradient.
Args: Args:
name (str): The name/alias of the variable, see :ref:`api_guide_Name` name (str): The name/alias of the Tensor, see :ref:`api_guide_Name`
for more details. for more details.
shape (list|tuple): List|Tuple of integers declaring the shape. You can shape (list|tuple): List|Tuple of integers declaring the shape. You can
set "None" at a dimension to indicate the dimension can be of any set "None" at a dimension to indicate the dimension can be of any
...@@ -54,7 +54,7 @@ def data_layer_not_check(name, shape, dtype='float32', lod_level=0): ...@@ -54,7 +54,7 @@ def data_layer_not_check(name, shape, dtype='float32', lod_level=0):
use LoD level, see :ref:`user_guide_lod_tensor` . Default: 0 use LoD level, see :ref:`user_guide_lod_tensor` . Default: 0
Returns: Returns:
Variable: The global variable that gives access to the data. Tensor: The global Tensor that gives access to the data.
""" """
helper = LayerHelper('data', **locals()) helper = LayerHelper('data', **locals())
shape = list(shape) shape = list(shape)
...@@ -87,7 +87,8 @@ def create_static_variable_gast_node(name): ...@@ -87,7 +87,8 @@ def create_static_variable_gast_node(name):
def create_fill_constant_node(name, value): def create_fill_constant_node(name, value):
func_code = "{} = paddle.fluid.layers.fill_constant(shape=[1], ".format(name) func_code = "{} = paddle.fluid.layers.fill_constant(shape=[1], ".format(
name)
if isinstance(value, bool): if isinstance(value, bool):
func_code += "dtype='bool', value={})".format(value) func_code += "dtype='bool', value={})".format(value)
return gast.parse(func_code).body[0] return gast.parse(func_code).body[0]
...@@ -110,7 +111,7 @@ def create_fill_constant_node(name, value): ...@@ -110,7 +111,7 @@ def create_fill_constant_node(name, value):
def to_static_variable(x): def to_static_variable(x):
''' '''
Translate a Python variable to PaddlePaddle static graph variable Translate a Python Tensor to PaddlePaddle static graph Tensor
''' '''
if isinstance(x, bool): if isinstance(x, bool):
return fill_constant(shape=[1], dtype='bool', value=x) return fill_constant(shape=[1], dtype='bool', value=x)
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册