episode: 0 training return: tensor(2.8096e-07, device='cuda:0', grad_fn=<AddBackward0>)
episode: 1 training return: tensor(2.2181e-07, device='cuda:0', grad_fn=<AddBackward0>)
episode: 2 training return: tensor(2.7333e-07, device='cuda:0', grad_fn=<AddBackward0>)
episode: 3 training return: tensor(3.4009e-07, device='cuda:0', grad_fn=<AddBackward0>)
epoch: 1 test_true_pfm: 0
episode: 4 training return: tensor(1.6024e-07, device='cuda:0', grad_fn=<AddBackward0>)
episode: 5 training return: tensor(2.2190e-07, device='cuda:0', grad_fn=<AddBackward0>)
episode: 6 training return: tensor(4.1637e-07, device='cuda:0', grad_fn=<AddBackward0>)
episode: 7 training return: tensor(3.0470e-07, device='cuda:0', grad_fn=<AddBackward0>)
epoch: 2 test_true_pfm: 0
episode: 8 training return: tensor(3.6797e-07, device='cuda:0', grad_fn=<AddBackward0>)
episode: 9 training return: tensor(6.1245e-07, device='cuda:0', grad_fn=<AddBackward0>)
episode: 10 training return: tensor(5.4984e-07, device='cuda:0', grad_fn=<AddBackward0>)
episode: 11 training return: tensor(1.6081e-07, device='cuda:0', grad_fn=<AddBackward0>)
epoch: 3 test_true_pfm: 0
episode: 12 training return: tensor(4.3546e-07, device='cuda:0', grad_fn=<AddBackward0>)
episode: 13 training return: tensor(2.5670e-07, device='cuda:0', grad_fn=<AddBackward0>)
episode: 14 training return: tensor(1.9255e-07, device='cuda:0', grad_fn=<AddBackward0>)
episode: 15 training return: tensor(3.3348e-07, device='cuda:0', grad_fn=<AddBackward0>)
epoch: 4 test_true_pfm: 0
episode: 16 training return: tensor(1.3055e-07, device='cuda:0', grad_fn=<AddBackward0>)
episode: 17 training return: tensor(1.9215e-07, device='cuda:0', grad_fn=<AddBackward0>)
episode: 18 training return: tensor(3.1404e-07, device='cuda:0', grad_fn=<AddBackward0>)
episode: 19 training return: tensor(3.4455e-07, device='cuda:0', grad_fn=<AddBackward0>)
epoch: 5 test_true_pfm: 0
episode: 20 training return: tensor(2.3420e-07, device='cuda:0', grad_fn=<AddBackward0>)
episode: 21 training return: tensor(1.6595e-07, device='cuda:0', grad_fn=<AddBackward0>)
episode: 22 training return: tensor(3.7291e-07, device='cuda:0', grad_fn=<AddBackward0>)
episode: 23 training return: tensor(2.0034e-07, device='cuda:0', grad_fn=<AddBackward0>)
epoch: 6 test_true_pfm: 0
episode: 24 training return: tensor(2.5965e-07, device='cuda:0', grad_fn=<AddBackward0>)
episode: 25 training return: tensor(3.1950e-07, device='cuda:0', grad_fn=<AddBackward0>)
