Initial commit

README.md

@@ -16,7 +16,7 @@ model-index:
       type: SpaceInvadersNoFrameskip-v4
     metrics:
     - type: mean_reward
-      value: 208.00 +/- 115.94
+      value: 329.00 +/- 157.97
       name: mean_reward
       verified: false
 ---
@@ -62,7 +62,7 @@ python -m rl_zoo3.push_to_hub --algo dqn --env SpaceInvadersNoFrameskip-v4 -f lo
 
 ## Hyperparameters
 ```python
-OrderedDict([('batch_size', 64),
+OrderedDict([('batch_size', 16),
              ('buffer_size', 1000),
              ('env_wrapper',
               ['stable_baselines3.common.atari_wrappers.AtariWrapper']),
@@ -70,12 +70,12 @@ OrderedDict([('batch_size', 64),
              ('exploration_fraction', 0.1),
              ('frame_stack', 4),
              ('gradient_steps', 1),
-             ('learning_rate', 0.001),
+             ('learning_rate', 0.01),
              ('learning_starts', 1000),
-             ('n_timesteps', 100000.0),
+             ('n_timesteps', 500000),
              ('optimize_memory_usage', False),
              ('policy', 'CnnPolicy'),
-             ('target_update_interval', 100),
+             ('target_update_interval', 500),
              ('train_freq', 4),
              ('normalize', False)])
 ```

args.yml

@@ -54,7 +54,7 @@
   - - save_replay_buffer
     - false
   - - seed
-    - 3371467990
+    - 736452160
   - - storage
     - null
   - - study_name

config.yml

@@ -1,6 +1,6 @@
 !!python/object/apply:collections.OrderedDict
 - - - batch_size
-    - 64
+    - 16
   - - buffer_size
     - 1000
   - - env_wrapper
@@ -14,16 +14,16 @@
   - - gradient_steps
     - 1
   - - learning_rate
-    - 0.001
+    - 0.01
   - - learning_starts
     - 1000
   - - n_timesteps
-    - 100000.0
+    - 500000
   - - optimize_memory_usage
     - false
   - - policy
     - CnnPolicy
   - - target_update_interval
-    - 100
+    - 500
   - - train_freq
     - 4

dqn-SpaceInvadersNoFrameskip-v4.zip

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2651c989dd180a94e1c53f6a57241a8b870d39e42e02c49daa1448035d414648
-size 27220128
+oid sha256:3700f3135d477a16e7fa23996c58ae8545e04f81620016102bf8b86ee0bf235e
+size 27220130

dqn-SpaceInvadersNoFrameskip-v4/policy.optimizer.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b20cb1f351a8096712cb98910d31d74ad7e6a32c095d9d5f3092062384dfcc95
+oid sha256:b3d16f90aa28fb5be9d9631a1a7119ad668809910d7295eec4119762a5aca9ab
 size 13506236

dqn-SpaceInvadersNoFrameskip-v4/policy.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d563d83241f9f14e6c4c5e78c4b8a2a84b07484290ff0f8645fe09e983c722ee
+oid sha256:3240abcbbbca24655dbbd504ade02454201ce84404a5e4217c47b95bcc77d37f
 size 13505370

replay.mp4

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:47cd452e4b4ebf8b4b28f0767f4fe1c0f7e93ae799ead95c269dbbabf3f30311
-size 278302
+oid sha256:d1220fd2663a829192a2b709c767eb2debf32246e6f8d3b5838a5840975ff8d0
+size 242623

results.json

@@ -1 +1 @@
-{"mean_reward": 208.0, "std_reward": 115.93532679903913, "is_deterministic": false, "n_eval_episodes": 10, "eval_datetime": "2024-02-06T17:59:29.284502"}
+{"mean_reward": 329.0, "std_reward": 157.96835126062436, "is_deterministic": false, "n_eval_episodes": 10, "eval_datetime": "2024-02-06T18:48:01.505333"}

train_eval_metrics.zip

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:aec4d32171ed47a765c21e7ef0ffbac1f2736f1f6616cbc77962bb222842826b
-size 5196
+oid sha256:7a16d9e30ad0abeb5cf3c28c64050a28c3623dec1f8286c1d34cb8842f0a2600
+size 20955