Constant¶

class paddle.nn.initializer. Constant ( value=0.0 ) [source]

Implement the constant initializer.

Parameters

value (float32|float64, optional) – constant value to initialize the parameter. Default: 0.0.

Examples

import paddle
import paddle.nn as nn

data = paddle.rand([30, 10, 2], dtype='float32')
linear = nn.Linear(2,
                    4,
                    weight_attr=nn.initializer.Constant(value=2.0))
res = linear(data)
print(linear.weight)
# Tensor(shape=[2, 4], dtype=float32, place=Place(gpu:0), stop_gradient=False,
#        [[2., 2., 2., 2.],
#         [2., 2., 2., 2.]])