shufflenet_v2_swish

paddle.vision.models. shufflenet_v2_swish ( pretrained=False, **kwargs ) [源代码]

使用 swish 进行激活的 ShuffleNetV2 模型,来自论文 "ShuffleNet V2: Practical Guidelines for Ecient CNN Architecture Design"

参数

  • pretrained (bool,可选) - 是否加载在imagenet数据集上的预训练权重。默认值:False。

返回

shufflenet_v2_swish模型,Layer的实例。

代码示例

import paddle
from paddle.vision.models import shufflenet_v2_swish

# build model
model = shufflenet_v2_swish()

# build model and load imagenet pretrained weight
# model = shufflenet_v2_swish(pretrained=True)

x = paddle.rand([1, 3, 224, 224])
out = model(x)

print(out.shape)