mindscience.models.layers.DropPath

class mindscience.models.layers.DropPath(dropout_rate=0.0)[source]

Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks).

Parameters: dropout_rate (float, optional) – The drop rate for the DropPath layer, greater than 0 and less equal than 1. Default: 0.0.

Inputs:

x (Tensor) - The input tensor.

Outputs:

output (Tensor) - The output tensor with drop path applied during training, or the original input during evaluation.

Examples

>>> import mindspore as ms
>>> from mindspore import Tensor
>>> import mindspore.common.dtype as mstype
>>> import numpy as np
>>> from mindscience.models.layers import DropPath
>>> ms.set_context(mode=ms.GRAPH_MODE, save_graphs=False)
>>> x = Tensor(np.ones([2, 3, 4]), mstype.float32)
>>> droppath = DropPath(dropout_rate=0.1)
>>> output = droppath(x)
>>> print(output.shape)
(2, 3, 4)