pushing model

README.md

@@ -16,7 +16,7 @@ model-index:
       type: Breakout-v5
     metrics:
     - type: mean_reward
-      value: 1.10 +/- 1.37
+      value: 679.80 +/- 208.17
       name: mean_reward
       verified: false
 ---
@@ -46,7 +46,7 @@ curl -OL https://huggingface.co/vwxyzjn/Breakout-v5-cleanba_ppo_envpool_impala_a
 curl -OL https://huggingface.co/vwxyzjn/Breakout-v5-cleanba_ppo_envpool_impala_atari_wrapper_naturecnn-seed1/raw/main/pyproject.toml
 curl -OL https://huggingface.co/vwxyzjn/Breakout-v5-cleanba_ppo_envpool_impala_atari_wrapper_naturecnn-seed1/raw/main/poetry.lock
 poetry install --all-extras
-python cleanba_ppo_envpool_impala_atari_wrapper_naturecnn.py --local-num-envs 120 --async-batch-size 40 --learner-device-ids 1 2 3 --total-timesteps 200000 --distributed --track --save-model --upload-model --env-id Breakout-v5 --seed 1
+python cleanba_ppo_envpool_impala_atari_wrapper_naturecnn.py --distributed --learner-device-ids 1 --track --save-model --upload-model --env-id Breakout-v5 --seed 1
 ```
 
 # Hyperparameters
@@ -54,7 +54,7 @@ python cleanba_ppo_envpool_impala_atari_wrapper_naturecnn.py --local-num-envs 12
 {'actor_device_ids': [0],
  'actor_devices': ['gpu:0'],
  'anneal_lr': True,
- 'async_batch_size': 40,
+ 'async_batch_size': 20,
  'async_update': 3,
  'batch_size': 15360,
  'capture_video': False,
@@ -66,14 +66,14 @@ python cleanba_ppo_envpool_impala_atari_wrapper_naturecnn.py --local-num-envs 12
  'exp_name': 'cleanba_ppo_envpool_impala_atari_wrapper_naturecnn',
  'gae_lambda': 0.95,
  'gamma': 0.99,
- 'global_learner_decices': ['gpu:1', 'gpu:2', 'gpu:3'],
+ 'global_learner_decices': ['gpu:1', 'gpu:3'],
  'hf_entity': '',
- 'learner_device_ids': [1, 2, 3],
- 'learner_devices': ['gpu:1', 'gpu:2', 'gpu:3'],
+ 'learner_device_ids': [1],
+ 'learner_devices': ['gpu:1'],
  'learning_rate': 0.00025,
- 'local_batch_size': 15360,
- 'local_minibatch_size': 3840,
- 'local_num_envs': 120,
+ 'local_batch_size': 7680,
+ 'local_minibatch_size': 1920,
+ 'local_num_envs': 60,
  'local_rank': 0,
  'max_grad_norm': 0.5,
  'minibatch_size': 3840,
@@ -81,20 +81,20 @@ python cleanba_ppo_envpool_impala_atari_wrapper_naturecnn.py --local-num-envs 12
  'num_envs': 120,
  'num_minibatches': 4,
  'num_steps': 128,
- 'num_updates': 13,
+ 'num_updates': 3255,
  'profile': False,
  'save_model': True,
  'seed': 1,
  'target_kl': None,
  'test_actor_learner_throughput': False,
  'torch_deterministic': True,
- 'total_timesteps': 200000,
+ 'total_timesteps': 50000000,
  'track': True,
  'update_epochs': 4,
  'upload_model': True,
  'vf_coef': 0.5,
  'wandb_entity': None,
  'wandb_project_name': 'cleanRL',
- 'world_size': 1}
+ 'world_size': 2}
 ```
     

cleanba_ppo_envpool_impala_atari_wrapper_naturecnn.cleanrl_model

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:56712db1863472d8852f66d9b7971cd340c9ed21ad9c1c9925245ddcfb88d03e
-size 6748012
+oid sha256:9a54c3a00b863c60d60fc8d6c2c3cca5c4c49b9b6b95931d9479bfec2bc7e3f6
+size 6747984

events.out.tfevents.1677099833.ip-26-0-136-122

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:66d528aeae85e6eebb5ed4b5ffbebd397284baf70eb765e14c20747e6d629183
+size 4754155