initialize tensor variables for bias terms b1 = torch.randn((1, ... loss = loss_function(output, target) valid_loss.append(loss.item())
確定! 回上一頁