modify cli; fix model position bug

sotopia/benchmark/cli.py

@@ -54,6 +54,7 @@ def check_existing_episodes(
             assert isinstance(episode, EpisodeLog), "episode should be an EpisodeLog"
             if episode.agents == agent_ids and episode.models == list(models.values()):
                 return True
+        breakpoint()
         return False
     else:
         return False
@@ -70,7 +71,6 @@ def initilize_benchmark_combo(dataset) -> list[EnvAgentComboStorage]:  # type: i
 
 
 def get_avg_reward(episodes: list[EpisodeLog], model_name: str) -> dict[str, float]:
-    breakpoint()
     rewards_list = []
     avg_reward_dict = {}
     for episode in episodes:
@@ -157,6 +157,7 @@ def run_async_benchmark_in_batch(
         model_names=model_names, tag=tag, env_agent_combo_storage_list=benchmark_combo
     )
     env_agent_combo_iter_length = sum(1 for _ in env_agent_combo_iter)
+    breakpoint()
     env_agent_combo_iter = _iterate_all_env_agent_combo_not_in_db(
         model_names=model_names, tag=tag, env_agent_combo_storage_list=benchmark_combo
     )
@@ -220,7 +221,7 @@ def run_async_benchmark_in_batch(
             number_of_fix_turns += 1
             if env_agent_combo_iter_length == 0 or number_of_fix_turns >= 5:
                 rewards_dict = get_avg_reward(simulated_episodes, model_names["agent2"])  # type: ignore
-                rewards_dict["model_name"] = model_names["agent2"]  # type: ignore
+                rewards_dict["model_name"] = model_names["agent1"]  # type: ignore
                 print(rewards_dict)
                 return
 
@@ -265,15 +266,15 @@ def cli(
     """A simple command-line interface example."""
     _set_up_logs(print_logs=print_logs)
     typer.echo(
-        f"Running benchmark for {model} and {partner_model} on task {task} with {evaluator_model} as the evaluator."
+        f"Running benchmark for {model} chatting with {partner_model} on task {task} with {evaluator_model} as the evaluator."
     )
     model = cast(LLM_Name, model)
     partner_model = cast(LLM_Name, partner_model)
     evaluator_model = cast(LLM_Name, evaluator_model)
     tag = f"benchmark_{model}_q"
     run_async_benchmark_in_batch(
         batch_size=batch_size,
-        model_names={"agent1": model, "agent2": partner_model, "env": evaluator_model},
+        model_names={"env": evaluator_model, "agent1": model, "agent2": partner_model},
         tag=tag,
         verbose=False,
     )