pytorch conv2d 值无法转换为 uint8_t 类型而不溢出答案

【问题标题】：pytorch conv2d value cannot be converted to type uint8_t without overflowpytorch conv2d 值无法转换为 uint8_t 类型而不溢出
【发布时间】：2020-05-31 20:29:58
【问题描述】：

我将带有dtype 的torch.uint8 的torch.Tensor 传递给nn.Conv2d 模块，它给出了错误

RuntimeError: value 不能被转换为类型 uint8_t 没有溢出：-0.0344873

我的 conv2d 定义为self.conv1 = nn.Conv2d(3, 6, 5)。当我将张量传递给self.conv1(x) 之类的模块时，错误出现在我的forward 方法中。张量具有形状 (4, 3, 480, 640)。我不确定如何解决这个问题。这是堆栈跟踪

Traceback (most recent call last):

  File "cnn.py", line 54, in <module>

    outputs = net(inputs)

  File "/Users/my_repos/venv_projc/lib/python3.7/site-packages/torch/nn/modules/module.py", line 532, in __call__

    result = self.forward(*input, **kwargs)

  File "cnn.py", line 24, in forward

    test = self.conv1(x)

  File "/Users/my_repos/venv_projc/lib/python3.7/site-packages/torch/nn/modules/module.py", line 532, in __call__

    result = self.forward(*input, **kwargs)

  File "/Users/my_repos/venv_projc/lib/python3.7/site-packages/torch/nn/modules/conv.py", line 345, in forward

    return self.conv2d_forward(input, self.weight)

  File "/Users/my_repos/venv_projc/lib/python3.7/site-packages/torch/nn/modules/conv.py", line 342, in conv2d_forward

    self.padding, self.dilation, self.groups)

RuntimeError: value cannot be converted to type uint8_t without overflow: -0.0344873

【问题讨论】：

标签： python pytorch conv-neural-network

【解决方案1】：

将张量转换为浮点数似乎可以修复它self.conv1(x.float())

【讨论】：