Views
No views yet
mean_reward = 217.95 +/- 51.27 over 10 episodes (greedy policy).LunarLander-actor.pt — policy network state dict (8 -> 64 -> 64 -> 64 -> 4, Tanh)LunarLander-critic.pt — value network state dict (8 -> 64 -> 64 -> 64 -> 1, Tanh)replay.mp4 — replay video of the greedy agentresults.json — evaluation results1policy = Policy(obs_dim=8, hidden_dim=64, act_dim=4, n_layers=2)
2policy.load_state_dict(torch.load("LunarLander-actor.pt"))
3action, _, _ = policy.act(torch.tensor(obs), evaluation=True)1env_id = LunarLander-v3
2p_hidden_dim = 64
3p_n_layers = 2
4value_hidden_dim = 64
5value_n_layers = 2
6gamma = 0.999
7total_timesteps = 2000000
8beta = 0.01
9num_envs = 16
10num_steps = 1024
11num_minibatch = 32
12update_epochs = 4
13learning_rate = 0.00025
14gae_lambda = 0.98
15eps = 0.2
16v_coeff = 0.5
17clip_value_loss = False