关于pytorch userwarning坑的记录一

记录一下最近使用pytorch时出现的几个关于userwarning的坑

1.UserWarning: masked_fill_ received a mask with dtype torch.uint8, this behavior is now deprecated,please use a mask with dtype torch.bool instead.

/opt/conda/condabld/pytorch_1587428091666/work/aten/src/ATen/native/cuda/LegacyDefinitions.cpp:19: UserWarning: masked_fill_ received a mask with dtype torch.uint8, this behavior is now deprecated,please use a mask with dtype torch.bool instead.
这种情况需要把`

masked_fill_(data) 

`
改成

masked_filled(data.bool())

2.RuntimeError: Subtraction, the - operator, with a bool tensor is not supported. If you are trying to invert a mask, use the ~ or logical_not() operator instead.

这是因为 pytorch版本的问题，出现这种问题一般是因为你跑的代码当初可能写在pytorch1.2版本之前，而你现在的运行环境>=pytorch1.2因此，汇报上述错误，这是只需把

1-mask（）

改成

~mask()

即可搞定