Initial commit

README.md

@@ -16,7 +16,7 @@ model-index:
       type: SpaceInvadersNoFrameskip-v4
     metrics:
     - type: mean_reward
-      value: 341.50 +/- 165.58
+      value: 655.50 +/- 152.90
       name: mean_reward
       verified: false
 ---
@@ -63,7 +63,7 @@ python -m rl_zoo3.push_to_hub --algo dqn --env SpaceInvadersNoFrameskip-v4 -f lo
 
 ## Hyperparameters
 ```python
-OrderedDict([('batch_size', 256),
+OrderedDict([('batch_size', 32),
              ('buffer_size', 100000),
              ('env_wrapper',
               ['stable_baselines3.common.atari_wrappers.AtariWrapper']),
@@ -71,9 +71,9 @@ OrderedDict([('batch_size', 256),
              ('exploration_fraction', 0.1),
              ('frame_stack', 4),
              ('gradient_steps', 1),
-             ('learning_rate', 'linear_0.001'),
+             ('learning_rate', 0.0001),
              ('learning_starts', 100000),
-             ('n_timesteps', 1000000),
+             ('n_timesteps', 1000000.0),
              ('optimize_memory_usage', False),
              ('policy', 'CnnPolicy'),
              ('target_update_interval', 1000),

args.yml

@@ -56,7 +56,7 @@
   - - save_replay_buffer
     - false
   - - seed
-    - 716030509
+    - 1815171918
   - - storage
     - null
   - - study_name

config.yml

@@ -1,6 +1,6 @@
 !!python/object/apply:collections.OrderedDict
 - - - batch_size
-    - 256
+    - 32
   - - buffer_size
     - 100000
   - - env_wrapper
@@ -14,11 +14,11 @@
   - - gradient_steps
     - 1
   - - learning_rate
-    - linear_0.001
+    - 0.0001
   - - learning_starts
     - 100000
   - - n_timesteps
-    - 1000000
+    - 1000000.0
   - - optimize_memory_usage
     - false
   - - policy

dqn-SpaceInvadersNoFrameskip-v4.zip

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:049cca6f33efbd9e42a84876087fd999f1a5fec01332f0d3ffc1afb60f3adfb9
-size 27221460
+oid sha256:51bf9b93faa09ca086c78e7fda2f7e220ca206a3ca081d8ac3a65e983c2903f8
+size 27221023

dqn-SpaceInvadersNoFrameskip-v4/policy.optimizer.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9456e51216364bbd3735d828afeff573d210a6a7b14e09ab904fb2f55f9a5c0c
+oid sha256:5c5791e91af8d54e6fdde0aeec7d41eee137dbfe0414c9b91c3a4a9c614087b5
 size 13506172

dqn-SpaceInvadersNoFrameskip-v4/policy.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fe9abd92efcd832f4537aa3a569f8aee87d14124ac56baf8be923cafec1bab31
+oid sha256:b745d4df2782d6bb8363cee968a6efe7dd027611f42059cfe731dc5e65a6923f
 size 13505370

results.json

@@ -1 +1 @@
-{"mean_reward": 341.5, "std_reward": 165.57551147437232, "is_deterministic": false, "n_eval_episodes": 10, "eval_datetime": "2025-02-15T18:15:21.826436"}
+{"mean_reward": 655.5, "std_reward": 152.8962066239709, "is_deterministic": false, "n_eval_episodes": 10, "eval_datetime": "2025-02-15T19:36:47.432441"}

train_eval_metrics.zip

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3302e792d66a2a2c1019f32d4d9fc8b8fff00899c1c1b6164b7c53042531276c
-size 43334
+oid sha256:d86665aa9044ef74c25a87cdcbd450198ba27f95554a23111cab3b3d51ff08d7
+size 37042