support async in agents (#1178)

* Make auto reply method pluggable

* support async

* async

* allow richer trigger types

* test list

* rename key

flaml/autogen/agentchat/agent.py

@@ -27,16 +27,23 @@ def name(self):
  def send(self, message: Union[Dict, str], recipient: "Agent", request_reply: Optional[bool] = None):
  """(Aabstract method) Send a message to another agent."""
 
+ async def a_send(self, message: Union[Dict, str], recipient: "Agent", request_reply: Optional[bool] = None):
+ """(Aabstract async method) Send a message to another agent."""
+
  def receive(self, message: Union[Dict, str], sender: "Agent", request_reply: Optional[bool] = None):
  """(Abstract method) Receive a message from another agent."""
 
+ async def a_receive(self, message: Union[Dict, str], sender: "Agent", request_reply: Optional[bool] = None):
+ """(Abstract async method) Receive a message from another agent."""
+
  def reset(self):
  """(Abstract method) Reset the agent."""
 
  def generate_reply(
  self,
  messages: Optional[List[Dict]] = None,
  sender: Optional["Agent"] = None,
+ **kwargs,
  ) -> Union[str, Dict, None]:
  """(Abstract method) Generate a reply based on the received messages.
 
@@ -46,3 +53,18 @@ def generate_reply(
  Returns:
  str or dict or None: the generated reply. If None, no reply is generated.
  """
+
+ async def a_generate_reply(
+ self,
+ messages: Optional[List[Dict]] = None,
+ sender: Optional["Agent"] = None,
+ **kwargs,
+ ) -> Union[str, Dict, None]:
+ """(Abstract async method) Generate a reply based on the received messages.
+
+ Args:
+ messages (list[dict]): a list of messages received.
+ sender: sender of an Agent instance.
+ Returns:
+ str or dict or None: the generated reply. If None, no reply is generated.
+ """

flaml/autogen/agentchat/responsive_agent.py

@@ -1,3 +1,4 @@
+import asyncio
 from collections import defaultdict
 import copy
 import json
@@ -309,6 +310,48 @@ def send(self, message: Union[Dict, str], recipient: Agent, request_reply: Optio
  "Message can't be converted into a valid ChatCompletion message. Either content or function_call must be provided."
  )
 
+ async def a_send(self, message: Union[Dict, str], recipient: Agent, request_reply: Optional[bool] = None) -> bool:
+ """(async) Send a message to another agent.
+
+ Args:
+ message (dict or str): message to be sent.
+ The message could contain the following fields (either content or function_call must be provided):
+ - content (str): the content of the message.
+ - function_call (str): the name of the function to be called.
+ - name (str): the name of the function to be called.
+ - role (str): the role of the message, any role that is not "function"
+ will be modified to "assistant".
+ - context (dict): the context of the message, which will be passed to
+ [autogen.Completion.create](../oai/Completion#create).
+ For example, one agent can send a message A as:
+ ```python
+ {
+ "content": lambda context: context["use_tool_msg"],
+ "context": {
+ "use_tool_msg": "Use tool X if they are relevant."
+ }
+ }
+ ```
+ Next time, one agent can send a message B with a different "use_tool_msg".
+ Then the content of message A will be refreshed to the new "use_tool_msg".
+ So effectively, this provides a way for an agent to send a "link" and modify
+ the content of the "link" later.
+ recipient (Agent): the recipient of the message.
+ request_reply (bool or None): whether to request a reply from the recipient.
+
+ Raises:
+ ValueError: if the message can't be converted into a valid ChatCompletion message.
+ """
+ # When the agent composes and sends the message, the role of the message is "assistant"
+ # unless it's "function".
+ valid = self._append_oai_message(message, "assistant", recipient)
+ if valid:
+ await recipient.a_receive(message, self, request_reply)
+ else:
+ raise ValueError(
+ "Message can't be converted into a valid ChatCompletion message. Either content or function_call must be provided."
+ )
+
  def _print_received_message(self, message: Union[Dict, str], sender: Agent):
  # print the message received
  print(colored(sender.name, "yellow"), "(to", f"{self.name}):\n", flush=True)
@@ -339,6 +382,16 @@ def _print_received_message(self, message: Union[Dict, str], sender: Agent):
  print(colored("*" * len(func_print), "green"), flush=True)
  print("\n", "-" * 80, flush=True, sep="")
 
+ def _process_received_message(self, message, sender):
+ message = self._message_to_dict(message)
+ # When the agent receives a message, the role of the message is "user". (If 'role' exists and is 'function', it will remain unchanged.)
+ valid = self._append_oai_message(message, "user", sender)
+ if not valid:
+ raise ValueError(
+ "Received message can't be converted into a valid ChatCompletion message. Either content or function_call must be provided."
+ )
+ self._print_received_message(message, sender)
+
  def receive(self, message: Union[Dict, str], sender: Agent, request_reply: Optional[bool] = None):
  """Receive a message from another agent.
 
@@ -361,20 +414,50 @@ def receive(self, message: Union[Dict, str], sender: Agent, request_reply: Optio
  Raises:
  ValueError: if the message can't be converted into a valid ChatCompletion message.
  """
- message = self._message_to_dict(message)
- # When the agent receives a message, the role of the message is "user". (If 'role' exists and is 'function', it will remain unchanged.)
- valid = self._append_oai_message(message, "user", sender)
- if not valid:
- raise ValueError(
- "Received message can't be converted into a valid ChatCompletion message. Either content or function_call must be provided."
- )
- self._print_received_message(message, sender)
+ self._process_received_message(message, sender)
  if request_reply is False or request_reply is None and self.reply_at_receive[sender] is False:
  return
  reply = self.generate_reply(sender=sender)
  if reply is not None:
  self.send(reply, sender)
 
+ async def a_receive(self, message: Union[Dict, str], sender: Agent, request_reply: Optional[bool] = None):
+ """(async) Receive a message from another agent.
+
+ Once a message is received, this function sends a reply to the sender or stop.
+ The reply can be generated automatically or entered manually by a human.
+
+ Args:
+ message (dict or str): message from the sender. If the type is dict, it may contain the following reserved fields (either content or function_call need to be provided).
+ 1. "content": content of the message, can be None.
+ 2. "function_call": a dictionary containing the function name and arguments.
+ 3. "role": role of the message, can be "assistant", "user", "function".
+ This field is only needed to distinguish between "function" or "assistant"/"user".
+ 4. "name": In most cases, this field is not needed. When the role is "function", this field is needed to indicate the function name.
+ 5. "context" (dict): the context of the message, which will be passed to
+ [autogen.Completion.create](../oai/Completion#create).
+ sender: sender of an Agent instance.
+ request_reply (bool or None): whether a reply is requested from the sender.
+ If None, the value is determined by `self.reply_at_receive[sender]`.
+
+ Raises:
+ ValueError: if the message can't be converted into a valid ChatCompletion message.
+ """
+ self._process_received_message(message, sender)
+ if request_reply is False or request_reply is None and self.reply_at_receive[sender] is False:
+ return
+ reply = await self.a_generate_reply(sender=sender)
+ if reply is not None:
+ await self.a_send(reply, sender)
+
+ def _prepare_chat(self, recipient, clear_history):
+ self.reset_consecutive_auto_reply_counter(recipient)
+ recipient.reset_consecutive_auto_reply_counter(self)
+ self.reply_at_receive[recipient] = recipient.reply_at_receive[self] = True
+ if clear_history:
+ self.clear_history(recipient)
+ recipient.clear_history(self)
+
  def initiate_chat(self, recipient: "ResponsiveAgent", clear_history: Optional[bool] = True, **context):
  """Initiate a chat with the recipient agent.
 
@@ -388,14 +471,25 @@ def initiate_chat(self, recipient: "ResponsiveAgent", clear_history: Optional[bo
  **context: any context information.
  "message" needs to be provided if the `generate_init_message` method is not overridden.
  """
- self.reset_consecutive_auto_reply_counter(recipient)
- recipient.reset_consecutive_auto_reply_counter(self)
- self.reply_at_receive[recipient] = recipient.reply_at_receive[self] = True
- if clear_history:
- self.clear_history(recipient)
- recipient.clear_history(self)
+ self._prepare_chat(recipient, clear_history)
  self.send(self.generate_init_message(**context), recipient)
 
+ async def a_initiate_chat(self, recipient: "ResponsiveAgent", clear_history: Optional[bool] = True, **context):
+ """(async) Initiate a chat with the recipient agent.
+
+ Reset the consecutive auto reply counter.
+ If `clear_history` is True, the chat history with the recipient agent will be cleared.
+ `generate_init_message` is called to generate the initial message for the agent.
+
+ Args:
+ recipient: the recipient agent.
+ clear_history (bool): whether to clear the chat history with the agent.
+ **context: any context information.
+ "message" needs to be provided if the `generate_init_message` method is not overridden.
+ """
+ self._prepare_chat(recipient, clear_history)
+ await self.a_send(self.generate_init_message(**context), recipient)
+
  def reset(self):
  """Reset the agent."""
  self.clear_history()
@@ -573,12 +667,12 @@ def generate_reply(
  """Reply based on the conversation history and the sender.
 
  Either messages or sender must be provided.
- Use registered class-specific reply functions to generate replies.
+ Use registered auto reply functions to generate replies.
  By default, the following functions are checked in order:
- 1. _check_termination_and_human_reply
- 2. _generate_function_call_reply
- 3. _generate_code_execution_reply
- 4. _generate_oai_reply
+ 1. check_termination_and_human_reply
+ 2. generate_function_call_reply
+ 3. generate_code_execution_reply
+ 4. generate_oai_reply
  Every function returns a tuple (final, reply).
  When a function returns final=False, the next function will be checked.
  So by default, termination and human reply will be checked first.
@@ -597,16 +691,68 @@ def generate_reply(
  assert messages is not None or sender is not None, "Either messages or sender must be provided."
  if sender is not None:
  for reply_func_tuple in self._reply_func_list:
- if exclude and reply_func_tuple["reply_func"] in exclude:
+ reply_func = reply_func_tuple["reply_func"]
+ if exclude and reply_func in exclude:
+ continue
+ if asyncio.coroutines.iscoroutinefunction(reply_func):
  continue
  if self._match_trigger(reply_func_tuple["trigger"], sender):
- final, reply = reply_func_tuple["reply_func"](
+ final, reply = reply_func(
  self, messages=messages, sender=sender, context=reply_func_tuple["context"]
  )
  if final:
  return reply
  return self._default_auto_reply
 
+ async def a_generate_reply(
+ self,
+ messages: Optional[List[Dict]] = None,
+ sender: Optional[Agent] = None,
+ exclude: Optional[List[Callable]] = None,
+ ) -> Union[str, Dict, None]:
+ """(async) Reply based on the conversation history and the sender.
+
+ Either messages or sender must be provided.
+ Use registered auto reply functions to generate replies.
+ By default, the following functions are checked in order:
+ 1. check_termination_and_human_reply
+ 2. generate_function_call_reply
+ 3. generate_code_execution_reply
+ 4. generate_oai_reply
+ Every function returns a tuple (final, reply).
+ When a function returns final=False, the next function will be checked.
+ So by default, termination and human reply will be checked first.
+ If not terminating and human reply is skipped, execute function or code and return the result.
+ AI replies are generated only when no code execution is performed.
+
+ Args:
+ messages: a list of messages in the conversation history.
+ default_reply (str or dict): default reply.
+ sender: sender of an Agent instance.
+ exclude: a list of functions to exclude.
+
+ Returns:
+ str or dict or None: reply. None if no reply is generated.
+ """
+ assert messages is not None or sender is not None, "Either messages or sender must be provided."
+ if sender is not None:
+ for reply_func_tuple in self._reply_func_list:
+ reply_func = reply_func_tuple["reply_func"]
+ if exclude and reply_func in exclude:
+ continue
+ if self._match_trigger(reply_func_tuple["trigger"], sender):
+ if asyncio.coroutines.iscoroutinefunction(reply_func):
+ final, reply = await reply_func(
+ self, messages=messages, sender=sender, context=reply_func_tuple["context"]
+ )
+ else:
+ final, reply = reply_func(
+ self, messages=messages, sender=sender, context=reply_func_tuple["context"]
+ )
+ if final:
+ return reply
+ return self._default_auto_reply
+
  def _match_trigger(self, trigger, sender):
  """Check if the sender matches the trigger."""
  if isinstance(trigger, str):

flaml/autogen/agentchat/user_proxy_agent.py

@@ -9,7 +9,6 @@ class UserProxyAgent(ResponsiveAgent):
  and `llm_config` to False. By default, the agent will prompt for human input every time a message is received.
  Code execution is enabled by default. LLM-based auto reply is disabled by default.
  To modify auto reply, register a method with (`register_auto_reply`)[responsive_agent#register_auto_reply].
- The method should have a similar signature with `_generate_oai_reply` method.
  To modify the way to get human input, override `get_human_input` method.
  To modify the way to execute code blocks, single code block, or function call, override `execute_code_blocks`,
  `run_code`, and `execute_function` methods respectively.

flaml/version.py

@@ -1 +1 @@
-__version__ = "2.0.0rc4"
+__version__ = "2.0.0rc5"

notebook/autogen_agentchat_auto_feedback_from_code_execution.ipynb

@@ -19,8 +19,8 @@
  "source": [
  "# Auto Generated Agent Chat: Task Solving with Code Generation, Execution & Debugging\n",
  "\n",
- "FLAML offers an experimental feature of interactive LLM agents, which can be used to solve various tasks with human or automatic feedback, including tasks that require using tools via code.\n",
- "Please find documentation about this feature [here](https://microsoft.github.io/FLAML/docs/Use-Cases/Auto-Generation#agents-experimental).\n",
+ "FLAML offers conversable LLM agents, which can be used to solve various tasks with human or automatic feedback, including tasks that require using tools via code.\n",
+ "Please find documentation about this feature [here](https://microsoft.github.io/FLAML/docs/Use-Cases/Auto-Generation#agents).\n",
  "\n",
  "In this notebook, we demonstrate how to use `AssistantAgent` and `UserProxyAgent` to write code and execute the code. Here `AssistantAgent` is an LLM-based agent that can write Python code (in a Python coding block) for a user to execute for a given task. `UserProxyAgent` is an agent which serves as a proxy for the human user to execute the code written by `AssistantAgent`, or automatically execute the code. Depending on the setting of `human_input_mode` and `max_consecutive_auto_reply`, the `UserProxyAgent` either solicits feedback from the human user or returns auto-feedback based on the result of code execution (success or failure and corresponding outputs) to `AssistantAgent`. `AssistantAgent` will debug the code and suggest new code if the result contains error. The two agents keep communicating to each other until the task is done.\n",
  "\n",

notebook/autogen_agentchat_chess.ipynb

@@ -32,7 +32,7 @@
  "outputs": [],
  "source": [
  "%%capture --no-stderr\n",
- "# %pip install flaml[autogen]~=2.0.0rc4\n",
+ "# %pip install flaml[autogen]~=2.0.0rc5\n",
  "%pip install chess -U"
  ]
  },

notebook/autogen_agentchat_function_call.ipynb

@@ -17,7 +17,7 @@
  "source": [
  "# Auto Generated Agent Chat: Task Solving with Provided Tools as Functions\n",
  "\n",
- "FLAML offers an experimental feature of interactive LLM agents, which can be used to solve various tasks with human or automatic feedback, including tasks that require using tools via code. Please find documentation about this feature [here](https://microsoft.github.io/FLAML/docs/Use-Cases/Auto-Generation#agents-experimental).\n",
+ "FLAML offers conversable LLM agents, which can be used to solve various tasks with human or automatic feedback, including tasks that require using tools via code. Please find documentation about this feature [here](https://microsoft.github.io/FLAML/docs/Use-Cases/Auto-Generation#agents).\n",
  "\n",
  "In this notebook, we demonstrate how to use `AssistantAgent` and `UserProxyAgent` to make function calls with the new feature of OpenAI models (in model version 0613). A specified prompt and function configs need to be passed to `AssistantAgent` to initialize the agent. The corresponding functions need to be passed to `UserProxyAgent`, which will be responsible for executing any function calls made by `AssistantAgent`. Besides this requirement of matching descriptions with functions, we recommend checking the system message in the `AssistantAgent` to make sure the instructions align with the function call descriptions.\n",
  "\n",

notebook/autogen_agentchat_groupchat.ipynb

@@ -32,7 +32,7 @@
  "outputs": [],
  "source": [
  "%%capture --no-stderr\n",
- "# %pip install flaml[autogen]~=2.0.0rc4"
+ "# %pip install flaml[autogen]~=2.0.0rc5"
  ]
  },
  {