Initial commit

README.md

@@ -16,7 +16,7 @@ model-index:
       type: SpaceInvadersNoFrameskip-v4
     metrics:
     - type: mean_reward
-      value: 339.50 +/- 235.12
+      value: 613.50 +/- 107.96
       name: mean_reward
       verified: false
 ---
@@ -57,17 +57,17 @@ python -m rl_zoo3.push_to_hub --algo dqn --env SpaceInvadersNoFrameskip-v4 -f lo
 
 ## Hyperparameters
 ```python
-OrderedDict([('batch_size', 64),
-             ('buffer_size', 50000),
+OrderedDict([('batch_size', 32),
+             ('buffer_size', 10000),
              ('env_wrapper',
               ['stable_baselines3.common.atari_wrappers.AtariWrapper']),
-             ('exploration_final_eps', 0.01),
-             ('exploration_fraction', 0.4),
+             ('exploration_final_eps', 0.005),
+             ('exploration_fraction', 0.025),
              ('frame_stack', 4),
              ('gradient_steps', 1),
-             ('learning_rate', 0.0003),
-             ('learning_starts', 10000),
-             ('n_timesteps', 1000000.0),
+             ('learning_rate', 1e-05),
+             ('learning_starts', 500),
+             ('n_timesteps', 150000.0),
              ('optimize_memory_usage', False),
              ('policy', 'CnnPolicy'),
              ('target_update_interval', 1000),

args.yml

@@ -18,7 +18,7 @@
   - - hyperparams
     - null
   - - log_folder
-    - logs/
+    - logs5/
   - - log_interval
     - -1
   - - max_total_trials
@@ -54,7 +54,7 @@
   - - save_replay_buffer
     - false
   - - seed
-    - 632072827
+    - 3569262184
   - - storage
     - null
   - - study_name
@@ -64,7 +64,7 @@
   - - track
     - false
   - - trained_agent
-    - ''
+    - logs4/dqn/SpaceInvadersNoFrameskip-v4_1/best_model.zip
   - - truncate_last_trajectory
     - true
   - - uuid

config.yml

@@ -1,24 +1,24 @@
 !!python/object/apply:collections.OrderedDict
 - - - batch_size
-    - 64
+    - 32
   - - buffer_size
-    - 50000
+    - 10000
   - - env_wrapper
     - - stable_baselines3.common.atari_wrappers.AtariWrapper
   - - exploration_final_eps
-    - 0.01
+    - 0.005
   - - exploration_fraction
-    - 0.4
+    - 0.025
   - - frame_stack
     - 4
   - - gradient_steps
     - 1
   - - learning_rate
-    - 0.0003
+    - 1.0e-05
   - - learning_starts
-    - 10000
+    - 500
   - - n_timesteps
-    - 1000000.0
+    - 150000.0
   - - optimize_memory_usage
     - false
   - - policy

dqn-SpaceInvadersNoFrameskip-v4.zip

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ee61dbf51cea4a8e30f1415f4dcc77eca3d3f86c60a85ca7349e9317d56cebca
-size 27224787
+oid sha256:59a2f6a282fd844c7c992fecca295f2d78865b5d56f5321f63083844c188bc4e
+size 27224786

dqn-SpaceInvadersNoFrameskip-v4/policy.optimizer.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fb224c1c760f125652d4a4dcb062b985afbdf59c79111432baf3385782f4ac0f
+oid sha256:c3080e2eaa429dc4e96740322089494dd76d0a49328351dafc6ea87460450ec1
 size 13505739

dqn-SpaceInvadersNoFrameskip-v4/policy.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ffef05e0114ceda2c005f2ea4154edfb15a6d6c28a496104e27db7b6a10e2519
+oid sha256:51cfe6a3dbb4e448627fc0f102bdce2344fb162ea346fa031d242b478ad0f15e
 size 13504937

replay.mp4

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a8464ae8bf454c0495aa7455114408da0f74b21cf5f6d0fcf73ab23c6d556a8a
-size 252815
+oid sha256:d35a278f857febde301d5aa4af87c71472ec39373325932911796b0e5fac40d7
+size 187047

results.json

@@ -1 +1 @@
-{"mean_reward": 339.5, "std_reward": 235.12177695823925, "is_deterministic": false, "n_eval_episodes": 10, "eval_datetime": "2022-12-21T10:54:59.964068"}
+{"mean_reward": 613.5, "std_reward": 107.95948314066717, "is_deterministic": false, "n_eval_episodes": 10, "eval_datetime": "2022-12-22T09:55:38.039851"}

train_eval_metrics.zip

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8f092a53d15b163656698228bcbed6f8e42a5ec37cf26af3235bf493c8eb248f
-size 38415
+oid sha256:78220371993493400f9b4d15fe2e769c1264a728c1972822cd112f7d1dedbca2
+size 5585