pushing model

README.md

@@ -16,7 +16,7 @@ model-index:
       type: Swimmer-v4
     metrics:
     - type: mean_reward
-      value: 124.70 +/- 1.56
+      value: 124.05 +/- 2.58
       name: mean_reward
       verified: false
 ---
@@ -46,7 +46,7 @@ curl -OL https://huggingface.co/sdpkjc/Swimmer-v4-ppo_fix_continuous_action-seed
 curl -OL https://huggingface.co/sdpkjc/Swimmer-v4-ppo_fix_continuous_action-seed1/raw/main/pyproject.toml
 curl -OL https://huggingface.co/sdpkjc/Swimmer-v4-ppo_fix_continuous_action-seed1/raw/main/poetry.lock
 poetry install --all-extras
-python ppo_fix_continuous_action.py --save-model --hf-entity sdpkjc --seed 1 --upload-model --track --env-id Swimmer-v4
+python ppo_fix_continuous_action.py --save-model --upload-model --hf-entity sdpkjc --env-id Swimmer-v4 --seed 1 --track
 ```
 
 # Hyperparameters

events.out.tfevents.1700813954.4090-171.1716227.0 → events.out.tfevents.1704457676.4090-171.127219.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5f5a2e881d890ddec39b87458115a8a1ff9d2c97eb90579eff8e19c390ddfb09
+oid sha256:6a6923c2e8c876db8656fa0ff57440c0edfe325041e98489c0396ce72f6aeb2c
 size 376390

ppo_fix_continuous_action.py

@@ -430,7 +430,7 @@ if __name__ == "__main__":
 
             # https://github.com/DLR-RM/stable-baselines3/pull/658
             for idx, trunc in enumerate(truncations):
-                if trunc:
+                if trunc and not terminations[idx]:
                     real_next_obs = infos["final_observation"][idx]
                     with torch.no_grad():
                         terminal_value = agent.get_value(torch.Tensor(real_next_obs).to(device)).reshape(1, -1)[0][0]