Initial commit

README.md

@@ -16,7 +16,7 @@ model-index:
       type: SpaceInvadersNoFrameskip-v4
     metrics:
     - type: mean_reward
-      value: 159.50 +/- 78.94
+      value: 280.00 +/- 102.10
       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', 64),
+OrderedDict([('batch_size', 32),
              ('buffer_size', 100000),
              ('env_wrapper',
               ['stable_baselines3.common.atari_wrappers.AtariWrapper']),
              ('exploration_final_eps', 0.01),
-             ('exploration_fraction', 0.15),
+             ('exploration_fraction', 0.11),
              ('frame_stack', 4),
              ('gradient_steps', 1),
-             ('learning_rate', 0.0001),
+             ('learning_rate', 0.001),
              ('learning_starts', 10000),
-             ('n_timesteps', 100000),
+             ('n_timesteps', 100000.0),
              ('optimize_memory_usage', False),
              ('policy', 'CnnPolicy'),
-             ('target_update_interval', 1000),
+             ('target_update_interval', 500),
              ('train_freq', 4),
              ('normalize', False)])
 ```

args.yml

@@ -54,7 +54,7 @@
   - - save_replay_buffer
     - false
   - - seed
-    - 1128719579
+    - 3055323658
   - - storage
     - null
   - - study_name

config.yml

@@ -1,6 +1,6 @@
 !!python/object/apply:collections.OrderedDict
 - - - batch_size
-    - 64
+    - 32
   - - buffer_size
     - 100000
   - - env_wrapper
@@ -8,22 +8,22 @@
   - - exploration_final_eps
     - 0.01
   - - exploration_fraction
-    - 0.15
+    - 0.11
   - - frame_stack
     - 4
   - - gradient_steps
     - 1
   - - learning_rate
-    - 0.0001
+    - 0.001
   - - learning_starts
     - 10000
   - - n_timesteps
-    - 100000
+    - 100000.0
   - - optimize_memory_usage
     - false
   - - policy
     - CnnPolicy
   - - target_update_interval
-    - 1000
+    - 500
   - - train_freq
     - 4

dqn-SpaceInvadersNoFrameskip-v4.zip

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e22cd686f6a19d503ddf9567ded67e933868bf57933425465ce54e434fecd6c3
-size 27220131
+oid sha256:3f0cc648a2923d0858303d0e5774d838b4c2d0500e564c298e72c69a8578ad56
+size 27220130

dqn-SpaceInvadersNoFrameskip-v4/policy.optimizer.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:691e18de355d24433631945fafe1eef516ffe2a403f048497682b625842bc00c
+oid sha256:38c38333e15f332e65df74e182e4ff5cfbc7f76835e9a20640eee0d8692d034b
 size 13506236

dqn-SpaceInvadersNoFrameskip-v4/policy.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8768b963fefab8fc818f05c98716f455359a47ac41a0a75480ef80a8f1e3e52b
+oid sha256:c0838eda037b3383303ecc344d9176419bc80daedde5d0ce543c884f2803af20
 size 13505370

replay.mp4

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c76c0272e96762db5a99183326edf73fb00324b6e1c3580d2f114e2b2fe384f2
-size 254671
+oid sha256:ff5e946cc231dbe64e0c1beab938cd430c84999f8e075125322efd5f6806f850
+size 268534

results.json

@@ -1 +1 @@
-{"mean_reward": 159.5, "std_reward": 78.94460082868238, "is_deterministic": false, "n_eval_episodes": 10, "eval_datetime": "2024-02-06T01:18:58.679650"}
+{"mean_reward": 280.0, "std_reward": 102.10288928331069, "is_deterministic": false, "n_eval_episodes": 10, "eval_datetime": "2024-02-06T01:34:34.492176"}

train_eval_metrics.zip

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:06d6535919fd30500f8b7f787da58f540079dc6953dfb6b3f209c402eda6f158
-size 5052
+oid sha256:70b899b4b7ccda4e36dc26dac005c255ae208d81b3c62591c2486029f36509d2
+size 4812