p-christ/Deep-Reinforcement-Learning-Algorithms-with-PyTorch

Is there any error in discriminator loss in GAIL?

Joywanglulu opened this issue · 0 comments

I'm wondering if the loss of discriminator in GAIL could be
"discrim_loss = discrim_criterion(g_o, zeros((states.shape[0], 1), device=device)) +
discrim_criterion(e_o, ones((expert_traj.shape[0], 1), device=device))"

instead of
"discrim_loss = discrim_criterion(g_o, ones((states.shape[0], 1), device=device)) +
discrim_criterion(e_o, zeros((expert_traj.shape[0], 1), device=device))"?