import torch
class test_cat(torch.nn.Module):
    def __init__(self) ->None:
        super().__init__()
        self.ones = torch.ones(1,2,4)
    def forward(self,x):
        y =torch.cat([x[:,:,1,:],self.ones], dim=-1)
        return y
input = torch. zeros(1,2,3,4)
model = test_cat()
trace = torch.jit.trace(model,input)
trace.save("test_cat.pt")

网友的代码，我就不明白，执行py后，为何会有tensor向量输出？请大佬指点一二，谢谢
--
FROM 120.242.238.*