From 3aa56d458ebe6254e6cca631c226f6d4a266ddb2 Mon Sep 17 00:00:00 2001
From: yitong <2969413251@qq.com>
Date: Wed, 25 Oct 2023 19:24:03 +0800
Subject: [PATCH] fix: make to_message asynchronous to accelerate compressing
 chat history of multiple agents

---
 agentverse/agents/tasksolving_agent/critic.py |  2 +-
 .../agents/tasksolving_agent/evaluator.py     | 20 ++++--
 .../agents/tasksolving_agent/executor.py      | 69 +------------------
 .../agents/tasksolving_agent/role_assigner.py | 14 ++--
 agentverse/agents/tasksolving_agent/solver.py | 14 ++--
 .../environments/tasksolving_env/basic.py     |  4 +-
 .../tasksolving_env/rules/base.py             |  8 +--
 .../rules/decision_maker/brainstorming.py     |  2 +-
 .../rules/decision_maker/central.py           |  2 +-
 .../rules/decision_maker/concurrent.py        |  2 +-
 .../rules/decision_maker/dynamic.py           |  4 +-
 .../rules/decision_maker/horizontal.py        |  2 +-
 .../rules/decision_maker/horizontal_tool.py   |  4 +-
 .../rules/decision_maker/vertical.py          |  2 +-
 .../decision_maker/vertical_solver_first.py   |  4 +-
 .../tasksolving_env/rules/evaluator/base.py   |  2 +-
 .../tasksolving_env/rules/evaluator/basic.py  |  8 +--
 .../rules/executor/code_test.py               | 30 +-------
 .../rules/executor/tool_using.py              |  4 +-
 .../rules/role_assigner/base.py               |  4 +-
 .../rules/role_assigner/role_description.py   |  8 +--
 agentverse/memory/chat_history.py             | 20 +++---
 22 files changed, 75 insertions(+), 154 deletions(-)

diff --git a/agentverse/agents/tasksolving_agent/critic.py b/agentverse/agents/tasksolving_agent/critic.py
index ad0541c44..89850dd61 100644
--- a/agentverse/agents/tasksolving_agent/critic.py
+++ b/agentverse/agents/tasksolving_agent/critic.py
@@ -88,7 +88,7 @@ async def astep(
 
         max_send_token -= prompt_token
 
-        history = self.memory.to_messages(
+        history = await self.memory.to_messages(
             self.name,
             start_index=-self.max_history,
             max_send_token=max_send_token,
diff --git a/agentverse/agents/tasksolving_agent/evaluator.py b/agentverse/agents/tasksolving_agent/evaluator.py
index d9c1375e7..78275b492 100644
--- a/agentverse/agents/tasksolving_agent/evaluator.py
+++ b/agentverse/agents/tasksolving_agent/evaluator.py
@@ -26,6 +26,17 @@ def step(
         task_description: str,
         all_role_description: str,
     ) -> EvaluatorMessage:
+        pass
+        # return parsed_response
+
+    async def astep(
+        self,
+        solution: str,
+        result: str,
+        task_description: str,
+        all_role_description: str,
+    ) -> EvaluatorMessage:
+        """Asynchronous version of step"""
         logger.debug("", self.name, Fore.MAGENTA)
         prepend_prompt, append_prompt, prompt_token = self.get_all_prompts(
             solution=solution,
@@ -49,7 +60,7 @@ def step(
 
         max_send_token -= prompt_token
 
-        history = self.memory.to_messages(
+        history = await self.memory.to_messages(
             self.name,
             max_send_token=max_send_token,
             model=model_name,
@@ -57,7 +68,7 @@ def step(
         parsed_response = None
         for i in range(self.max_retry):
             try:
-                response = self.llm.generate_response(
+                response = await self.llm.agenerate_response(
                     prepend_prompt, history, append_prompt
                 )
                 parsed_response = self.output_parser.parse(response)
@@ -78,11 +89,6 @@ def step(
             advice=parsed_response[1] if parsed_response is not None else "",
         )
         return message
-        # return parsed_response
-
-    async def astep(self, solution: str) -> EvaluatorMessage:
-        """Asynchronous version of step"""
-        pass
 
     def _fill_prompt_template(self, solution: str, task_description: str) -> str:
         """Fill the placeholders in the prompt template
diff --git a/agentverse/agents/tasksolving_agent/executor.py b/agentverse/agents/tasksolving_agent/executor.py
index bd8b07c4c..6e4ebc4ee 100644
--- a/agentverse/agents/tasksolving_agent/executor.py
+++ b/agentverse/agents/tasksolving_agent/executor.py
@@ -23,72 +23,7 @@ class ExecutorAgent(BaseAgent):
     def step(
         self, task_description: str, solution: str, tools: List[dict] = [], **kwargs
     ) -> ExecutorMessage:
-        logger.debug("", self.name, Fore.MAGENTA)
-        prepend_prompt, append_prompt, prompt_token = self.get_all_prompts(
-            task_description=task_description,
-            solution=solution,
-            agent_name=self.name,
-            **kwargs,
-        )
-
-        model_name = self.llm.args.model
-
-        if model_name.startswith("gpt-3.5-turbo"):
-            tokens_per_message = 4
-        else:
-            tokens_per_message = 3
-
-        max_send_token = self.llm.send_token_limit(model_name)
-        if len(prepend_prompt) > 0:
-            max_send_token -= tokens_per_message
-        if (len(append_prompt)) > 0:
-            max_send_token -= tokens_per_message
-
-        max_send_token -= prompt_token
-
-        history = self.memory.to_messages(
-            self.name,
-            start_index=-self.max_history,
-            max_send_token=max_send_token,
-            model=model_name,
-        )
-        parsed_response = None
-        for i in range(self.max_retry):
-            try:
-                response = self.llm.generate_response(
-                    prepend_prompt, history, append_prompt, tools
-                )
-                parsed_response = self.output_parser.parse(response)
-                break
-            except (KeyboardInterrupt, bdb.BdbQuit):
-                raise
-            except Exception as e:
-                logger.error(e)
-                logger.warn("Retrying...")
-                continue
-
-        if parsed_response is None:
-            logger.error(f"{self.name} failed to generate valid response.")
-        if isinstance(parsed_response, AgentFinish):
-            message = ExecutorMessage(
-                content=parsed_response.return_values["output"],
-                sender=self.name,
-                sender_agent=self,
-            )
-        elif isinstance(parsed_response, AgentAction):
-            message = ExecutorMessage(
-                content=parsed_response.log,
-                sender=self.name,
-                sender_agent=self,
-                tool_name=parsed_response.tool,
-                tool_input=parsed_response.tool_input,
-            )
-        else:
-            raise ValueError(
-                f"Error response type: {type(parsed_response)}. Only support \
-                    AgentFinish and AgentAction. Modify your output parser."
-            )
-        return message
+        pass
 
     async def astep(
         self, task_description: str, solution: str, tools: List[dict] = [], **kwargs
@@ -116,7 +51,7 @@ async def astep(
 
         max_send_token -= prompt_token
 
-        history = self.memory.to_messages(
+        history = await self.memory.to_messages(
             self.name,
             start_index=-self.max_history,
             max_send_token=max_send_token,
diff --git a/agentverse/agents/tasksolving_agent/role_assigner.py b/agentverse/agents/tasksolving_agent/role_assigner.py
index ba752eee0..8fb3271a4 100644
--- a/agentverse/agents/tasksolving_agent/role_assigner.py
+++ b/agentverse/agents/tasksolving_agent/role_assigner.py
@@ -22,6 +22,12 @@ class RoleAssignerAgent(BaseAgent):
     def step(
         self, advice: str, task_description: str, cnt_critic_agents: int
     ) -> RoleAssignerMessage:
+        pass
+
+    async def astep(
+        self, advice: str, task_description: str, cnt_critic_agents: int
+    ) -> RoleAssignerMessage:
+        """Asynchronous version of step"""
         logger.debug("", self.name, Fore.MAGENTA)
         prepend_prompt, append_prompt, prompt_token = self.get_all_prompts(
             advice=advice,
@@ -44,13 +50,13 @@ def step(
 
         max_send_token -= prompt_token
 
-        history = self.memory.to_messages(
+        history = await self.memory.to_messages(
             self.name, max_send_token=max_send_token, model=model_name
         )
         parsed_response = None
         for i in range(self.max_retry):
             try:
-                response = self.llm.generate_response(
+                response = await self.llm.agenerate_response(
                     prepend_prompt, history, append_prompt
                 )
                 parsed_response = self.output_parser.parse(response)
@@ -76,10 +82,6 @@ def step(
         )
         return message
 
-    async def astep(self, env_description: str = "") -> RoleAssignerMessage:
-        """Asynchronous version of step"""
-        pass
-
     def _fill_prompt_template(
         self, advice, task_description: str, cnt_critic_agents: int
     ) -> str:
diff --git a/agentverse/agents/tasksolving_agent/solver.py b/agentverse/agents/tasksolving_agent/solver.py
index d65e397ae..dd95ea55f 100644
--- a/agentverse/agents/tasksolving_agent/solver.py
+++ b/agentverse/agents/tasksolving_agent/solver.py
@@ -25,6 +25,12 @@ class SolverAgent(BaseAgent):
     def step(
         self, former_solution: str, advice: str, task_description: str = "", **kwargs
     ) -> SolverMessage:
+        pass
+
+    async def astep(
+        self, former_solution: str, advice: str, task_description: str = "", **kwargs
+    ) -> SolverMessage:
+        """Asynchronous version of step"""
         logger.debug("", self.name, Fore.MAGENTA)
         # prompt = self._fill_prompt_template(
         #     former_solution, critic_opinions, advice, task_description
@@ -52,7 +58,7 @@ def step(
 
         max_send_token -= prompt_token
 
-        history = self.memory.to_messages(
+        history = await self.memory.to_messages(
             self.name,
             start_index=-self.max_history,
             max_send_token=max_send_token,
@@ -61,7 +67,7 @@ def step(
         parsed_response = None
         for i in range(self.max_retry):
             try:
-                response = self.llm.generate_response(
+                response = await self.llm.agenerate_response(
                     prepend_prompt, history, append_prompt
                 )
                 parsed_response = self.output_parser.parse(response)
@@ -85,10 +91,6 @@ def step(
         )
         return message
 
-    async def astep(self, env_description: str = "") -> SolverMessage:
-        """Asynchronous version of step"""
-        pass
-
     def _fill_prompt_template(
         self,
         former_solution: str,
diff --git a/agentverse/environments/tasksolving_env/basic.py b/agentverse/environments/tasksolving_env/basic.py
index 8e4631a24..457430311 100644
--- a/agentverse/environments/tasksolving_env/basic.py
+++ b/agentverse/environments/tasksolving_env/basic.py
@@ -51,7 +51,7 @@ async def step(
         logger.info(f"Loop Round {self.cnt_turn}")
 
         # ================== EXPERT RECRUITMENT ==================
-        agents = self.rule.role_assign(
+        agents = await self.rule.role_assign(
             self.task_description, self.agents, self.cnt_turn, advice
         )
         description = "\n".join([agent.role_description for agent in agents])
@@ -79,7 +79,7 @@ async def step(
         # ================== EXECUTION ==================
 
         # ================== EVALUATION ==================
-        score, advice = self.rule.evaluate(
+        score, advice = await self.rule.evaluate(
             self.task_description, self.agents, plan, result
         )
         logs.append(
diff --git a/agentverse/environments/tasksolving_env/rules/base.py b/agentverse/environments/tasksolving_env/rules/base.py
index 2e25c017c..a1fcb0dc9 100644
--- a/agentverse/environments/tasksolving_env/rules/base.py
+++ b/agentverse/environments/tasksolving_env/rules/base.py
@@ -68,7 +68,7 @@ def build_components(config: Dict, registry):
             **kwargs,
         )
 
-    def role_assign(
+    async def role_assign(
         self,
         task_description: str,
         agents: List[BaseAgent],
@@ -79,7 +79,7 @@ def role_assign(
         if self.role_assign_only_once and cnt_turn > 0:
             agents = [agents[AGENT_TYPES.SOLVER]] + agents[AGENT_TYPES.CRITIC]
         else:
-            agents = self.role_assigner.step(
+            agents = await self.role_assigner.astep(
                 role_assigner=agents[AGENT_TYPES.ROLE_ASSIGNMENT],
                 group_members=[agents[AGENT_TYPES.SOLVER]] + agents[AGENT_TYPES.CRITIC],
                 advice=advice,
@@ -137,7 +137,7 @@ async def execute(
             agents[AGENT_TYPES.SOLVER].add_message_to_memory(results)
         return results
 
-    def evaluate(
+    async def evaluate(
         self,
         task_description: str,
         agents: List[BaseAgent],
@@ -162,7 +162,7 @@ def evaluate(
         #         logger.error("Bad response from human evaluator!")
         #     return ([comprehensiveness, detailedness, feasibility, novelty], advice)
         # else:
-        evaluation = self.evaluator.step(
+        evaluation = await self.evaluator.astep(
             agent=agents[AGENT_TYPES.EVALUATION],
             solution=solution,
             result=result,
diff --git a/agentverse/environments/tasksolving_env/rules/decision_maker/brainstorming.py b/agentverse/environments/tasksolving_env/rules/decision_maker/brainstorming.py
index a6db1a5f6..39bd35ae9 100644
--- a/agentverse/environments/tasksolving_env/rules/decision_maker/brainstorming.py
+++ b/agentverse/environments/tasksolving_env/rules/decision_maker/brainstorming.py
@@ -53,7 +53,7 @@ async def astep(
                 Fore.YELLOW,
             )
 
-        result = agents[0].step(previous_plan, advice, task_description)
+        result = await agents[0].astep(previous_plan, advice, task_description)
         for agent in agents:
             agent.memory.reset()
         self.broadcast_messages(
diff --git a/agentverse/environments/tasksolving_env/rules/decision_maker/central.py b/agentverse/environments/tasksolving_env/rules/decision_maker/central.py
index 5d7bf5702..1d682e3da 100644
--- a/agentverse/environments/tasksolving_env/rules/decision_maker/central.py
+++ b/agentverse/environments/tasksolving_env/rules/decision_maker/central.py
@@ -47,7 +47,7 @@ async def astep(
             ),
         )
         agents[1].add_message_to_memory([result])
-        result = agents[0].step(
+        result = await agents[0].astep(
             previous_plan, advice, task_description, chat_record=result.content
         )
         return [result]
diff --git a/agentverse/environments/tasksolving_env/rules/decision_maker/concurrent.py b/agentverse/environments/tasksolving_env/rules/decision_maker/concurrent.py
index cc34e00b7..f3979b851 100644
--- a/agentverse/environments/tasksolving_env/rules/decision_maker/concurrent.py
+++ b/agentverse/environments/tasksolving_env/rules/decision_maker/concurrent.py
@@ -59,7 +59,7 @@ async def astep(
             last_reviews = nonempty_reviews
 
         agents[0].add_message_to_memory(last_reviews)
-        result = agents[0].step(previous_plan, advice, task_description)
+        result = await agents[0].astep(previous_plan, advice, task_description)
         # agents[0].add_message_to_memory([result])
         self.broadcast_messages(agents, [result])
         return [result]
diff --git a/agentverse/environments/tasksolving_env/rules/decision_maker/dynamic.py b/agentverse/environments/tasksolving_env/rules/decision_maker/dynamic.py
index d6b6d72fe..c1fccf923 100644
--- a/agentverse/environments/tasksolving_env/rules/decision_maker/dynamic.py
+++ b/agentverse/environments/tasksolving_env/rules/decision_maker/dynamic.py
@@ -56,7 +56,7 @@ async def astep(
             #    Fore.YELLOW,
             # )
 
-            previous_sentence = manager.step(
+            previous_sentence = await manager.astep(
                 previous_plan, review, advice, task_description, previous_sentence
             )
             reviews.append(previous_sentence)
@@ -76,7 +76,7 @@ async def astep(
                 nonempty_reviews.append(review)
         agents[0].add_message_to_memory(nonempty_reviews)
 
-        result = agents[0].step(previous_plan, advice, task_description)
+        result = await agents[0].astep(previous_plan, advice, task_description)
 
         return [result]
 
diff --git a/agentverse/environments/tasksolving_env/rules/decision_maker/horizontal.py b/agentverse/environments/tasksolving_env/rules/decision_maker/horizontal.py
index b2a8c5703..ef9f43108 100644
--- a/agentverse/environments/tasksolving_env/rules/decision_maker/horizontal.py
+++ b/agentverse/environments/tasksolving_env/rules/decision_maker/horizontal.py
@@ -50,7 +50,7 @@ async def astep(
                 Fore.YELLOW,
             )
 
-        result = agents[0].step(previous_plan, advice, task_description)
+        result = await agents[0].astep(previous_plan, advice, task_description)
         return [result]
 
     def reset(self):
diff --git a/agentverse/environments/tasksolving_env/rules/decision_maker/horizontal_tool.py b/agentverse/environments/tasksolving_env/rules/decision_maker/horizontal_tool.py
index 5cea85eab..b30e88046 100644
--- a/agentverse/environments/tasksolving_env/rules/decision_maker/horizontal_tool.py
+++ b/agentverse/environments/tasksolving_env/rules/decision_maker/horizontal_tool.py
@@ -77,7 +77,9 @@ async def astep(
             if end_flag:
                 break
 
-        result: SolverMessage = agents[0].step(previous_plan, advice, task_description)
+        result: SolverMessage = await agents[0].astep(
+            previous_plan, advice, task_description
+        )
         result_list = []
         for res in result.content:
             res_tmp = deepcopy(result)
diff --git a/agentverse/environments/tasksolving_env/rules/decision_maker/vertical.py b/agentverse/environments/tasksolving_env/rules/decision_maker/vertical.py
index d8adf594d..0bcc4688f 100644
--- a/agentverse/environments/tasksolving_env/rules/decision_maker/vertical.py
+++ b/agentverse/environments/tasksolving_env/rules/decision_maker/vertical.py
@@ -50,7 +50,7 @@ async def astep(
             if not review.is_agree and review.content != "":
                 nonempty_reviews.append(review)
         agents[0].add_message_to_memory(nonempty_reviews)
-        result = agents[0].step(previous_plan, advice, task_description)
+        result = await agents[0].astep(previous_plan, advice, task_description)
         agents[0].add_message_to_memory([result])
         return [result]
 
diff --git a/agentverse/environments/tasksolving_env/rules/decision_maker/vertical_solver_first.py b/agentverse/environments/tasksolving_env/rules/decision_maker/vertical_solver_first.py
index 97114f455..c3c38af8e 100644
--- a/agentverse/environments/tasksolving_env/rules/decision_maker/vertical_solver_first.py
+++ b/agentverse/environments/tasksolving_env/rules/decision_maker/vertical_solver_first.py
@@ -38,7 +38,7 @@ async def astep(
             self.broadcast_messages(
                 agents, [Message(content=advice, sender="Evaluator")]
             )
-        previous_plan = agents[0].step(previous_plan, advice, task_description)
+        previous_plan = await agents[0].astep(previous_plan, advice, task_description)
         self.broadcast_messages(agents, [previous_plan])
         logger.info("", f"Initial Plan:\n{previous_plan.content}", Fore.BLUE)
         for i in range(self.max_inner_turns):
@@ -65,7 +65,7 @@ async def astep(
                 logger.info("", "Consensus Reached!.", Fore.GREEN)
                 break
             self.broadcast_messages(agents, nonempty_reviews)
-            previous_plan = agents[0].step(previous_plan, advice, task_description)
+            previous_plan = await agents[0].astep(previous_plan, advice, task_description)
             logger.info("", f"Updated Plan:\n{previous_plan.content}", Fore.BLUE)
             self.broadcast_messages(agents, [previous_plan])
         result = previous_plan
diff --git a/agentverse/environments/tasksolving_env/rules/evaluator/base.py b/agentverse/environments/tasksolving_env/rules/evaluator/base.py
index f3d72ad98..de83528c1 100644
--- a/agentverse/environments/tasksolving_env/rules/evaluator/base.py
+++ b/agentverse/environments/tasksolving_env/rules/evaluator/base.py
@@ -20,7 +20,7 @@ class BaseEvaluator(BaseModel):
     """
 
     @abstractmethod
-    def step(
+    def astep(
         self,
         agent: EvaluatorAgent,
         solution: List[SolverMessage],
diff --git a/agentverse/environments/tasksolving_env/rules/evaluator/basic.py b/agentverse/environments/tasksolving_env/rules/evaluator/basic.py
index a7738fe21..847234f03 100644
--- a/agentverse/environments/tasksolving_env/rules/evaluator/basic.py
+++ b/agentverse/environments/tasksolving_env/rules/evaluator/basic.py
@@ -14,7 +14,7 @@
 class BasicEvaluator(BaseEvaluator):
     cnt_agents: int = 0
 
-    def step(
+    async def astep(
         self,
         agent: EvaluatorAgent,
         solution: List[SolverMessage],
@@ -27,7 +27,7 @@ def step(
         flatten_solution = "\n".join([s.content for s in solution])
         flatten_result = "\n".join([r.content for r in result])
         flatten_all_role_description = "\n".join(all_role_description)
-        evaluation = agent.step(
+        evaluation = await agent.astep(
             flatten_solution,
             flatten_result,
             task_description,
@@ -40,7 +40,7 @@ def step(
 class BasicEvaluator(BaseEvaluator):
     cnt_agents: int = 0
 
-    def step(
+    async def astep(
         self,
         agent: EvaluatorAgent,
         solution: List[SolverMessage],
@@ -54,7 +54,7 @@ def step(
         flatten_result = "\n".join([r.content for r in result])
         flatten_all_role_description = "\n".join(all_role_description)
         agent.add_message_to_memory(result)
-        evaluation = agent.step(
+        evaluation = await agent.astep(
             flatten_solution,
             flatten_result,
             task_description,
diff --git a/agentverse/environments/tasksolving_env/rules/executor/code_test.py b/agentverse/environments/tasksolving_env/rules/executor/code_test.py
index 121aabc67..1a60b720e 100644
--- a/agentverse/environments/tasksolving_env/rules/executor/code_test.py
+++ b/agentverse/environments/tasksolving_env/rules/executor/code_test.py
@@ -71,35 +71,7 @@ def step(
         *args,
         **kwargs,
     ) -> Any:
-        solution = solution[0].content
-        os.makedirs("tmp", exist_ok=True)
-        self.write_to_file("tmp/main.py", solution)
-        manager = multiprocessing.Manager()
-        result = manager.list()
-        if task_description not in self.has_test:
-            response = agent.step(task_description, solution).content
-            self.write_to_file(response["file_path"], response["code"])
-            self.has_test[task_description] = f"python {response['file_path']}"
-            p = multiprocessing.Process(
-                target=execute_command, args=(f"python {response['file_path']}", result)
-            )
-            p.start()
-            p.join(timeout=self.timeout + 1)
-            if p.is_alive():
-                p.kill()
-            # result = execute_command(f"python {response['file_path']}")
-        else:
-            # result = execute_command(self.has_test[task_description])
-            p = multiprocessing.Process(
-                target=execute_command, args=(self.has_test[task_description], result)
-            )
-            p.start()
-            p.join(timeout=self.timeout + 1)
-            if p.is_alive():
-                p.kill()
-        if not result:
-            result.append("Execution timed out.")
-        return [ExecutorMessage(content=result[0], sender="Code Tester")]
+        pass
 
     def write_to_file(self, file_name, file_content):
         # TODO: generalize this method to a common tool
diff --git a/agentverse/environments/tasksolving_env/rules/executor/tool_using.py b/agentverse/environments/tasksolving_env/rules/executor/tool_using.py
index 5f177fc1c..62bc30429 100644
--- a/agentverse/environments/tasksolving_env/rules/executor/tool_using.py
+++ b/agentverse/environments/tasksolving_env/rules/executor/tool_using.py
@@ -258,8 +258,8 @@ async def _summarize_webpage(webpage, question):
                         ],
                         function_call={"name": "parse_web_text"},
                     )
-                except e:
-                    logger.error(e)
+                except Exception as e:
+                    logger.error("Failed to call the tool. Exception: " + str(e))
                     continue
                 arguments = ast.literal_eval(
                     JsonRepair(
diff --git a/agentverse/environments/tasksolving_env/rules/role_assigner/base.py b/agentverse/environments/tasksolving_env/rules/role_assigner/base.py
index 726abf52a..22a3127c7 100644
--- a/agentverse/environments/tasksolving_env/rules/role_assigner/base.py
+++ b/agentverse/environments/tasksolving_env/rules/role_assigner/base.py
@@ -19,7 +19,7 @@ class BaseRoleAssigner(BaseModel):
     """
 
     @abstractmethod
-    def step(
+    def astep(
         self,
         role_assigner: RoleAssignerAgent,
         group_members: List[CriticAgent],
@@ -40,7 +40,7 @@ class DummyRoleAssigner(BaseRoleAssigner):
     The base class of role assignment class.
     """
 
-    def step(
+    def astep(
         self,
         role_assigner: RoleAssignerAgent,
         group_members: List[CriticAgent],
diff --git a/agentverse/environments/tasksolving_env/rules/role_assigner/role_description.py b/agentverse/environments/tasksolving_env/rules/role_assigner/role_description.py
index 1d7490c83..b28b64408 100644
--- a/agentverse/environments/tasksolving_env/rules/role_assigner/role_description.py
+++ b/agentverse/environments/tasksolving_env/rules/role_assigner/role_description.py
@@ -16,7 +16,7 @@ class DescriptionAssigner(BaseRoleAssigner):
     Generates descriptions for each agent.
     """
 
-    def step(
+    async def astep(
         self,
         role_assigner: RoleAssignerAgent,
         group_members: List[CriticAgent],
@@ -28,7 +28,7 @@ def step(
         assert task_description != ""
         assert len(group_members) > 0
 
-        roles = role_assigner.step(advice, task_description, len(group_members))
+        roles = await role_assigner.astep(advice, task_description, len(group_members))
         if len(roles.content) < len(group_members):
             raise ValueError(
                 f"Number of roles ({len(roles.content)}) and number of group members ({len(group_members)}) do not match."
@@ -50,7 +50,7 @@ class DescriptionNameAssigner(BaseRoleAssigner):
     Generates description and name for each agent.
     """
 
-    def step(
+    async def astep(
         self,
         role_assigner: RoleAssignerAgent,
         group_members: List[CriticAgent],
@@ -63,7 +63,7 @@ def step(
         assert len(group_members) > 0
 
         # roles: [{'name': 'xxx', 'description': 'xxx'}, ...]
-        roles = role_assigner.step(advice, task_description, len(group_members))
+        roles = await role_assigner.astep(advice, task_description, len(group_members))
 
         if len(group_members) < 2:
             pass
diff --git a/agentverse/memory/chat_history.py b/agentverse/memory/chat_history.py
index f05624bf0..f2188c065 100644
--- a/agentverse/memory/chat_history.py
+++ b/agentverse/memory/chat_history.py
@@ -78,7 +78,7 @@ def to_string(self, add_sender_prefix: bool = False) -> str:
         else:
             return "\n".join([message.content for message in self.messages])
 
-    def to_messages(
+    async def to_messages(
         self,
         my_name: str = "",
         start_index: int = 0,
@@ -141,7 +141,9 @@ def to_messages(
                 prompt, messages, max_send_token, model
             )
             if trimmed_history:
-                new_summary_msg, _ = self.trim_messages(list(prompt), model, messages)
+                new_summary_msg, _ = await self.trim_messages(
+                    list(prompt), model, messages
+                )
                 prompt.append(new_summary_msg)
             messages = prompt
         return messages
@@ -149,7 +151,7 @@ def to_messages(
     def reset(self) -> None:
         self.messages = []
 
-    def trim_messages(
+    async def trim_messages(
         self, current_message_chain: list[dict], model: str, history: List[dict]
     ) -> tuple[dict, list[dict]]:
         new_messages_not_in_chain = [
@@ -159,7 +161,7 @@ def trim_messages(
         if not new_messages_not_in_chain:
             return self.summary_message(), []
 
-        new_summary_message = self.update_running_summary(
+        new_summary_message = await self.update_running_summary(
             new_events=new_messages_not_in_chain, model=model
         )
 
@@ -168,7 +170,7 @@ def trim_messages(
 
         return new_summary_message, new_messages_not_in_chain
 
-    def update_running_summary(
+    async def update_running_summary(
         self,
         new_events: list[Message],
         model: str = "gpt-3.5-turbo",
@@ -217,7 +219,7 @@ def update_running_summary(
                 batch_tlength + event_tlength
                 > max_input_tokens - prompt_template_length - summary_tlength
             ):
-                self._update_summary_with_batch(batch, model, max_summary_length)
+                await self._update_summary_with_batch(batch, model, max_summary_length)
                 summary_tlength = count_string_tokens(self.summary, model)
                 batch = [event]
                 batch_tlength = event_tlength
@@ -226,18 +228,18 @@ def update_running_summary(
                 batch_tlength += event_tlength
 
         if batch:
-            self._update_summary_with_batch(batch, model, max_summary_length)
+            await self._update_summary_with_batch(batch, model, max_summary_length)
 
         return self.summary_message()
 
-    def _update_summary_with_batch(
+    async def _update_summary_with_batch(
         self, new_events_batch: list[dict], model: str, max_summary_length: int
     ) -> None:
         prompt = self.SUMMARIZATION_PROMPT.format(
             summary=self.summary, new_events=new_events_batch
         )
 
-        self.summary = openai.ChatCompletion.create(
+        self.summary = await openai.ChatCompletion.acreate(
             messages=[{"role": "user", "content": prompt}],
             model=model,
             max_tokens=max_summary_length,