w = torch.tensor(torch.randn([3, 1]), requires_grad=True) ... f = torch.stack([x * x, x, torch.ones_like(x)], 1). yhat = torch.squeeze(f @ w ...
確定! 回上一頁