5dba182362d6d3b7d87fe852a4d707bd7ec3153a,main.py,,main,#,43

Before Change

        next_value = actor_critic(Variable(rollouts.observations[-1], volatile=True))[0].data

        if hasattr(actor_critic, "obs_filter"):
            actor_critic.obs_filter.update(rollouts.observations[:-1].view(-1, *obs_shape))

        rollouts.compute_returns(next_value, args.use_gae, args.gamma, args.tau)

After Change

        for i in range(args.num_processes)
    ])

    if len(envs.observation_space.shape) == 1:
        envs = VecNormalize(envs)

    obs_shape = envs.observation_space.shape
    obs_shape = (obs_shape[0] * args.num_stack, *obs_shape[1:])

Instances