尝试使用 torch.load(PATH) 加载预先训练的模型时获取 Unicode 解码错误

Question

尝试使用 torch.load(PATH( 加载 ResNet 18 预训练模型，但收到 Unicode 解码错误，请帮助。

Traceback (most recent call last):
File "main.py", line 312, in <module>
main()
File "main.py", line 138, in main
checkpoint = torch.load(args.resume)
File "F:InsSoftAnacondalibsite-packagestorchserialization.py", line 593, in load
return _legacy_load(opened_file, map_location, pickle_module, **pickle_load_args)
File "F:InsSoftAnacondalibsite-packagestorchserialization.py", line 773, in _legacy_load
result = unpickler.load()
UnicodeDecodeError: 'utf-8' codec can't decode byte 0xbe in position 2: invalid start byte

Answer 1

每当模型在火炬版本< 0.4上预训练但使用火炬版本> 0.4进行测试/<恢复时，就会发生此错误。>所以使用checkpoint = torch.load(args.resume,encoding='latin1')