LazyConvTranspose2d¶
- class torch.nn.LazyConvTranspose2d(out_channels, kernel_size, stride=1, padding=0, output_padding=0, groups=1, bias=True, dilation=1, padding_mode='zeros', device=None, dtype=None)[source]¶
一个
torch.nn.ConvTranspose2d
模块,它对in_channels
参数进行延迟初始化。in_channels
参数ConvTranspose2d
来自input.size(1)
推断。将延迟初始化的属性是 weight 和 bias。有关延迟模块及其局限性的更多文档,请查看
torch.nn.modules.lazy.LazyModuleMixin
。- 参数
- cls_to_become¶
ConvTranspose2d
的别名