Merge pull request #3 from Benjamin-eecs/fix/tictactoe

fix: update is_ppo flag
waterhorse1 · Dec 19, 2024 · 1b7d42b · 1b7d42b
2 parents b0e31eb + be11ea0
commit 1b7d42b
Showing 1 changed file with 1 addition and 1 deletion.
diff --git a/nlrl/policy/llm_policy.py b/nlrl/policy/llm_policy.py
@@ -26,7 +26,7 @@ def __init__(
         self.tp_size = model_tp_size
         self.sample_config = sample_config
         self.is_gpt4 = "gpt-4" in model_path.lower()
-        self.is_ppo = "ppo" in model_path.lower()
+        self.is_ppo = "-ppo" in model_path.lower()
         print(f"model_path: {model_path}, is_gpt4: {self.is_gpt4}, is_ppo: {self.is_ppo}")
 
         if self.is_gpt4: