Initial commit

README.md

@@ -16,7 +16,7 @@ model-index:
       type: SpaceInvadersNoFrameskip-v4
     metrics:
     - type: mean_reward
-      value: 1059.50 +/- 346.63
+      value: 424.00 +/- 249.38
       name: mean_reward
       verified: false
 ---
@@ -63,21 +63,24 @@ python -m rl_zoo3.push_to_hub --algo dqn --env SpaceInvadersNoFrameskip-v4 -f lo
 
 ## Hyperparameters
 ```python
-OrderedDict([('batch_size', 32),
+OrderedDict([('batch_size', 100),
              ('buffer_size', 100000),
              ('env_wrapper',
               ['stable_baselines3.common.atari_wrappers.AtariWrapper']),
-             ('exploration_final_eps', 0.01),
-             ('exploration_fraction', 0.1),
+             ('exploration_final_eps', 0.13071780683854003),
+             ('exploration_fraction', 0.4477029703130554),
              ('frame_stack', 4),
+             ('gamma', 0.9),
              ('gradient_steps', 1),
-             ('learning_rate', 0.0001),
-             ('learning_starts', 100000),
+             ('learning_rate', 0.00011447658254850165),
+             ('learning_starts', 20000),
              ('n_timesteps', 9000000.0),
+             ('net_arch', 'medium'),
              ('optimize_memory_usage', False),
              ('policy', 'CnnPolicy'),
+             ('subsample_steps', 2),
              ('target_update_interval', 1000),
-             ('train_freq', 4),
+             ('train_freq', 8),
              ('normalize', False)])
 ```
 

dqn-SpaceInvadersNoFrameskip-v4.zip

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:503e8d4d9484f272e4a98563a524c8dddcd8845c5d4b9b7ccdfd9f3716da546d
-size 27220781
+oid sha256:3737ee51769185ac1b5f0d20384c0c92dcc30e446056391b9cd875990d0fb69c
+size 27220878

replay.mp4

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:51287969ed8f20349eeb1df7a42b99d8cafa5f7f9aab85b0678b5ba91fe6ae6b
-size 316484
+oid sha256:205a7e5a1be28d7e8d33286c5d64922a122cdc492fb0f0f070b7fe2e16414d84
+size 348430

results.json

@@ -1 +1 @@
-{"mean_reward": 1059.5, "std_reward": 346.6298458009639, "is_deterministic": false, "n_eval_episodes": 10, "eval_datetime": "2024-12-06T20:23:24.373824"}
+{"mean_reward": 424.0, "std_reward": 249.3772243008571, "is_deterministic": false, "n_eval_episodes": 10, "eval_datetime": "2024-12-07T12:26:25.554944"}