Initial commit

README.md

@@ -16,7 +16,7 @@ model-index:
       type: SpaceInvadersNoFrameskip-v4
     metrics:
     - type: mean_reward
-      value: 434.50 +/- 316.96
+      value: 860.50 +/- 399.22
       name: mean_reward
       verified: false
 ---
@@ -62,20 +62,20 @@ python -m rl_zoo3.push_to_hub --algo dqn --env SpaceInvadersNoFrameskip-v4 -f lo
 
 ## Hyperparameters
 ```python
-OrderedDict([('batch_size', 32),
-             ('buffer_size', 100000),
+OrderedDict([('batch_size', 64),
+             ('buffer_size', 150000),
              ('env_wrapper',
               ['stable_baselines3.common.atari_wrappers.AtariWrapper']),
              ('exploration_final_eps', 0.01),
-             ('exploration_fraction', 0.1),
+             ('exploration_fraction', 0.2),
              ('frame_stack', 4),
              ('gradient_steps', 1),
              ('learning_rate', 0.0001),
              ('learning_starts', 100000),
-             ('n_timesteps', 1000000.0),
+             ('n_timesteps', 10000000.0),
              ('optimize_memory_usage', False),
              ('policy', 'CnnPolicy'),
-             ('target_update_interval', 1000),
+             ('target_update_interval', 5000),
              ('train_freq', 4),
              ('normalize', False)])
 ```

args.yml

@@ -56,7 +56,7 @@
   - - save_replay_buffer
     - false
   - - seed
-    - 1222019558
+    - 447592518
   - - storage
     - null
   - - study_name

config.yml

@@ -1,14 +1,14 @@
 !!python/object/apply:collections.OrderedDict
 - - - batch_size
-    - 32
+    - 64
   - - buffer_size
-    - 100000
+    - 150000
   - - env_wrapper
     - - stable_baselines3.common.atari_wrappers.AtariWrapper
   - - exploration_final_eps
     - 0.01
   - - exploration_fraction
-    - 0.1
+    - 0.2
   - - frame_stack
     - 4
   - - gradient_steps
@@ -18,12 +18,12 @@
   - - learning_starts
     - 100000
   - - n_timesteps
-    - 1000000.0
+    - 10000000.0
   - - optimize_memory_usage
     - false
   - - policy
     - CnnPolicy
   - - target_update_interval
-    - 1000
+    - 5000
   - - train_freq
     - 4

dqn-SpaceInvadersNoFrameskip-v4.zip

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7c48c297a9d713f825ef6fec265ac84e8b8e6a97e04f41bf7456097cffd227bb
-size 27220460
+oid sha256:5aae8ff3bf2de68460fe74151e65129c91b048025bdd549bbfdfbeeedf7d267e
+size 27220465

dqn-SpaceInvadersNoFrameskip-v4/policy.optimizer.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c467b1057e41943b89453064c4309a3d6db91641dd9d32c2a12eef7c51448b36
+oid sha256:f0dfec666d041505fd9befc5271616c1b6e35181609f12379fc71ea4009ac0c3
 size 13506236

dqn-SpaceInvadersNoFrameskip-v4/policy.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:779b82dd371b8495b0a87b32d4f211d627df1917e08cc1d89115b4b4115b8ce3
+oid sha256:4ad807988e9b0514bcd6c070b0134840b0f20724fa3f1e4985ec9ee2cdc171b4
 size 13505370

replay.mp4

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:22ce2b6b153e432b2736799603dd37871385d4fd3f4bc99213ff92c9858f2668
-size 237921
+oid sha256:aa0f3e87e3e50c1bfff22c620af9eeff0f1eaed43826be6b70e62cd98456aa1a
+size 222419

results.json

@@ -1 +1 @@
-{"mean_reward": 434.5, "std_reward": 316.957804762716, "is_deterministic": false, "n_eval_episodes": 10, "eval_datetime": "2024-01-17T13:55:17.263550"}
+{"mean_reward": 860.5, "std_reward": 399.22080356614686, "is_deterministic": false, "n_eval_episodes": 10, "eval_datetime": "2024-01-17T22:34:19.500060"}

train_eval_metrics.zip

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6cac45e899f034ef45f9892215196017c23ede4a0c5906715f1b27546c2f2597
-size 36248
+oid sha256:0f649f099f9575970ba115d26f53835f2b75031fc7bdfcc21ff77f5f4036b5c8
+size 281687