已解决: RuntimeError: Input type (torch.cuda.FloatTensor) and weight type (torch.cuda.HalfTensor) should be the same
Opened this issue · 3 comments
dorannn commented
在position coding时,输入固定float,在yolo训练val时会自动进行半精度训练,将position()输出类型转换成input就好
position(......).to(x.dtype)
Deargoodnight commented
哥们,问一哈,可以搞到自然语言处理用吗?
dorannn commented
哥们,问一哈,可以搞到自然语言处理用吗?
不知道呀,我是目标检测
JasonYangCode commented
Good Job!