yeyupiaoling/AudioClassification-Pytorch

RuntimeError: Input type (torch.FloatTensor) and weight type (torch.cuda.FloatTensor) should be the same or input should be a MKLDNN tensor and weight is a dense tensor

Bill-JLN opened this issue 3 years ago · 7 comments

Bill-JLN commented 3 years ago

我更改了输入为tensorfloat但是还是报相似的错误

yeyupiaoling commented 3 years ago

类型错误，你有改吗？没看到。你这边修改了，你拉下最新代码试试。

Bill-JLN commented 3 years ago

我等会把最新的代码拉一下改一下

yeyupiaoling commented 3 years ago

主要是这里：

AudioClassification-Pytorch/reader.py

Line 22 in 017d398

spec_mag = spec_mag.astype('float32')

AudioClassification-Pytorch/infer.py

Line 22 in 017d398

spec_mag = spec_mag.astype('float32')

👍2
❤️1

Bill-JLN commented 3 years ago

这两行改成float64吗

yeyupiaoling commented 3 years ago

这是改好的。就是float32

Bill-JLN commented 3 years ago

行，我试试

Bill-JLN commented 3 years ago

解决了，跪谢大佬

👍1