Algabri

I fixed this issue following [this](https://stackoverflow.com/a/68527319/11162983)

I fixed it by updating tensorflow ==2.2.0 to tensorflow ==2.3 and numpy-1.20.0

I changed this line: `grad_seq = [torch.ones(1).cuda(gpu) for _ in range(len(loss_seq))]` To be: `grad_seq = [torch.tensor(1, dtype=torch.float).cuda(gpu) for _ in range(len(loss_seq))]` It is working fine now.

I am facing the same problem