variable dimension issue

[salman-qadir]

File "e:\python\cod\codes\rl-pendulum\Pendulum_DQN\pendulum_dqn.py", line 155, in action, real_action, count = agent.choose_action(torch.FloatTensor(state)) ^^^^^^^^^^^^^^^^^^^^^^^^ ValueError: expected sequence of length 3 at dim 1 (got 0)

[Elstuhn]

To fix this, go to line 149 and change it from state = env.reset() to state, _ = env.reset() This code is according to old gym and does not account for new gym

You'll likely encounter another problem at line 156 as well, change any instance of state_prime, reward, done, _ = env.step([real_action]) to state_prime, reward, done, truncated, _ = env.step([real_action]) and then change at line 153 from while not done to while not (done or truncated)