mirror of
https://github.com/microsoft/autogen.git
synced 2025-10-31 09:50:11 +00:00
simplify the initiation of chat (#1131)
* simplify the initiation of chat * version update * include openai * completion
This commit is contained in:
Chi Wang
GitHub
parent
7665f73e4b
commit
16f0fcd6f8
@ -63,13 +63,13 @@ class Agent:
|
||||
oai_message["role"] = "function" if message.get("role") == "function" else role
|
||||
self._oai_conversations[conversation_id].append(oai_message)
|
||||
|
||||
def _send(self, message: Union[Dict, str], recipient):
|
||||
def send(self, message: Union[Dict, str], recipient):
|
||||
"""Send a message to another agent."""
|
||||
# When the agent composes and sends the message, the role of the message is "assistant". (If 'role' exists and is 'function', it will remain unchanged.)
|
||||
self._append_oai_message(message, "assistant", recipient.name)
|
||||
recipient.receive(message, self)
|
||||
|
||||
def _receive(self, message: Union[Dict, str], sender):
|
||||
def _receive(self, message: Union[Dict, str], sender: "Agent"):
|
||||
"""Receive a message from another agent.
|
||||
|
||||
Args:
|
||||
|
||||
@ -42,7 +42,7 @@ class AssistantAgent(Agent):
|
||||
|
||||
super().receive(message, sender)
|
||||
responses = oai.ChatCompletion.create(messages=self._oai_conversations[sender.name], **self._config)
|
||||
self._send(oai.ChatCompletion.extract_text_or_function_call(responses)[0], sender)
|
||||
self.send(oai.ChatCompletion.extract_text_or_function_call(responses)[0], sender)
|
||||
|
||||
def reset(self):
|
||||
self._sender_dict.clear()
|
||||
|
||||
@ -222,7 +222,7 @@ class MathUserProxyAgent(UserProxyAgent):
|
||||
self._previous_code = ""
|
||||
self.last_reply = None
|
||||
|
||||
def _execute_one_python_code(self, pycode):
|
||||
def execute_one_python_code(self, pycode):
|
||||
"""Execute python code blocks.
|
||||
|
||||
Previous python code will be saved and executed together with the new code.
|
||||
@ -278,7 +278,7 @@ class MathUserProxyAgent(UserProxyAgent):
|
||||
self._previous_code = tmp
|
||||
return output, is_success
|
||||
|
||||
def _execute_one_wolfram_query(self, query: str):
|
||||
def execute_one_wolfram_query(self, query: str):
|
||||
"""
|
||||
Run one wolfram query and return the output.
|
||||
return:
|
||||
@ -302,7 +302,7 @@ class MathUserProxyAgent(UserProxyAgent):
|
||||
# no code block is found, lang should be `UNKNOWN``
|
||||
if default_reply == "":
|
||||
default_reply = "Continue. Please keep solving the problem until you need to query. (If you get to the answer, put it in \\boxed{}.)"
|
||||
self._send(default_reply, sender)
|
||||
self.send(default_reply, sender)
|
||||
else:
|
||||
is_success, all_success = True, True
|
||||
reply = ""
|
||||
@ -311,9 +311,9 @@ class MathUserProxyAgent(UserProxyAgent):
|
||||
if not lang:
|
||||
lang = infer_lang(code)
|
||||
if lang == "python":
|
||||
output, is_success = self._execute_one_python_code(code)
|
||||
output, is_success = self.execute_one_python_code(code)
|
||||
elif lang == "wolfram":
|
||||
output, is_success = self._execute_one_wolfram_query(code)
|
||||
output, is_success = self.execute_one_wolfram_query(code)
|
||||
else:
|
||||
output = "Error: Unknown language."
|
||||
is_success = False
|
||||
@ -338,7 +338,7 @@ class MathUserProxyAgent(UserProxyAgent):
|
||||
self._accum_invalid_q_per_step = 0
|
||||
reply = "Please revisit the problem statement and your reasoning. If you think this step is correct, solve it yourself and continue the next step. Otherwise, correct this step."
|
||||
|
||||
self._send(reply, sender)
|
||||
self.send(reply, sender)
|
||||
|
||||
|
||||
# Imported from langchain. Langchain is licensed under MIT License:
|
||||
|
||||
@ -77,7 +77,7 @@ class UserProxyAgent(Agent):
|
||||
or str value of the docker image name to use."""
|
||||
return self._use_docker
|
||||
|
||||
def _execute_code(self, code_blocks):
|
||||
def execute_code(self, code_blocks):
|
||||
"""Execute the code and return the result."""
|
||||
logs_all = ""
|
||||
for code_block in code_blocks:
|
||||
@ -185,19 +185,19 @@ class UserProxyAgent(Agent):
|
||||
def auto_reply(self, message: dict, sender, default_reply=""):
|
||||
"""Generate an auto reply."""
|
||||
if "function_call" in message:
|
||||
is_exec_success, func_return = self._execute_function(message["function_call"])
|
||||
self._send(func_return, sender)
|
||||
_, func_return = self._execute_function(message["function_call"])
|
||||
self.send(func_return, sender)
|
||||
return
|
||||
|
||||
code_blocks = extract_code(message["content"])
|
||||
if len(code_blocks) == 1 and code_blocks[0][0] == UNKNOWN:
|
||||
# no code block is found, lang should be `UNKNOWN`
|
||||
self._send(default_reply, sender)
|
||||
self.send(default_reply, sender)
|
||||
else:
|
||||
# try to execute the code
|
||||
exitcode, logs = self._execute_code(code_blocks)
|
||||
exitcode, logs = self.execute_code(code_blocks)
|
||||
exitcode2str = "execution succeeded" if exitcode == 0 else "execution failed"
|
||||
self._send(f"exitcode: {exitcode} ({exitcode2str})\nCode output: {logs}", sender)
|
||||
self.send(f"exitcode: {exitcode} ({exitcode2str})\nCode output: {logs}", sender)
|
||||
|
||||
def receive(self, message: Union[Dict, str], sender):
|
||||
"""Receive a message from the sender agent.
|
||||
@ -230,9 +230,36 @@ class UserProxyAgent(Agent):
|
||||
if reply:
|
||||
# reset the consecutive_auto_reply_counter
|
||||
self._consecutive_auto_reply_counter[sender.name] = 0
|
||||
self._send(reply, sender)
|
||||
self.send(reply, sender)
|
||||
return
|
||||
|
||||
self._consecutive_auto_reply_counter[sender.name] += 1
|
||||
print("\n>>>>>>>> NO HUMAN INPUT RECEIVED. USING AUTO REPLY FOR THE USER...", flush=True)
|
||||
self.auto_reply(message, sender, default_reply=reply)
|
||||
|
||||
def generate_init_prompt(self, *args, **kwargs) -> Union[str, Dict]:
|
||||
"""Generate the initial prompt for the agent.
|
||||
|
||||
Override this function to customize the initial prompt based on user's request.
|
||||
"""
|
||||
return args[0]
|
||||
|
||||
def initiate_chat(self, recipient, *args, **kwargs):
|
||||
"""Initiate a chat with the receiver agent.
|
||||
|
||||
`generate_init_prompt` is called to generate the initial prompt for the agent.
|
||||
|
||||
Args:
|
||||
receiver: the receiver agent.
|
||||
*args: any additional arguments.
|
||||
**kwargs: any additional keyword arguments.
|
||||
"""
|
||||
self.send(self.generate_init_prompt(*args, **kwargs), recipient)
|
||||
|
||||
def register_function(self, function_map: Dict[str, Callable]):
|
||||
"""Register functions to the agent.
|
||||
|
||||
Args:
|
||||
function_map: a dictionary mapping function names to functions.
|
||||
"""
|
||||
self._function_map.update(function_map)
|
||||
|
||||
@ -31,7 +31,7 @@
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"# %pip install flaml[mathchat]==2.0.0rc2"
|
||||
"# %pip install flaml[mathchat]~=2.0.0rc4"
|
||||
]
|
||||
},
|
||||
{
|
||||
@ -122,14 +122,16 @@
|
||||
" system_message=\"You are a helpful assistant.\",\n",
|
||||
" request_timeout=600, \n",
|
||||
" seed=42, \n",
|
||||
" config_list=config_list)\n",
|
||||
" config_list=config_list,\n",
|
||||
")\n",
|
||||
"\n",
|
||||
"# 2. create the MathUserProxyAgent instance named \"mathproxyagent\"\n",
|
||||
"# By default, the human_input_mode is \"NEVER\", which means the agent will not ask for human input.\n",
|
||||
"mathproxyagent = MathUserProxyAgent(\n",
|
||||
" name=\"MathChatAgent\", \n",
|
||||
" human_input_mode=\"NEVER\",\n",
|
||||
" use_docker=False)"
|
||||
" use_docker=False,\n",
|
||||
")"
|
||||
]
|
||||
},
|
||||
{
|
||||
@ -283,11 +285,8 @@
|
||||
"# given a math problem, we use the mathproxyagent to generate a prompt to be sent to the assistant as the initial message.\n",
|
||||
"# the assistant receives the message and generates a response. The response will be sent back to the mathproxyagent for processing.\n",
|
||||
"# The conversation continues until the termination condition is met, in MathChat, the termination condition is the detect of \"\\boxed{}\" in the response.\n",
|
||||
"math_problem = \"Find all $x$ that satisfy the inequality $(2x+10)(x+3)<(3x+9)(x+8)$. Express your answer in interval notation.\"\n",
|
||||
"assistant.receive(\n",
|
||||
" message=mathproxyagent.generate_init_prompt(math_problem),\n",
|
||||
" sender=mathproxyagent,\n",
|
||||
")"
|
||||
"math_problem = \"Find all $x$ that satisfy the inequality $(2x+10)(x+3)<(3x+9)(x+8)$. Express your answer in interval notation.\"\n",
|
||||
"mathproxyagent.initiate_chat(assistant, math_problem)"
|
||||
]
|
||||
},
|
||||
{
|
||||
@ -429,11 +428,8 @@
|
||||
"source": [
|
||||
"assistant.reset()\n",
|
||||
"\n",
|
||||
"math_problem = \"For what negative value of $k$ is there exactly one solution to the system of equations \\\\begin{align*}\\ny &= 2x^2 + kx + 6 \\\\\\\\\\ny &= -x + 4?\\n\\\\end{align*}\"\n",
|
||||
"assistant.receive(\n",
|
||||
" mathproxyagent.generate_init_prompt(math_problem),\n",
|
||||
" mathproxyagent,\n",
|
||||
")"
|
||||
"math_problem = \"For what negative value of $k$ is there exactly one solution to the system of equations \\\\begin{align*}\\ny &= 2x^2 + kx + 6 \\\\\\\\\\ny &= -x + 4?\\n\\\\end{align*}\"\n",
|
||||
"mathproxyagent.initiate_chat(assistant, math_problem)"
|
||||
]
|
||||
},
|
||||
{
|
||||
@ -561,11 +557,8 @@
|
||||
"source": [
|
||||
"assistant.reset()\n",
|
||||
"\n",
|
||||
"math_problem = \"Find all positive integer values of $c$ such that the equation $x^2-7x+c=0$ only has roots that are real and rational. Express them in decreasing order, separated by commas.\"\n",
|
||||
"assistant.receive(\n",
|
||||
" mathproxyagent.generate_init_prompt(math_problem),\n",
|
||||
" mathproxyagent,\n",
|
||||
")"
|
||||
"math_problem = \"Find all positive integer values of $c$ such that the equation $x^2-7x+c=0$ only has roots that are real and rational. Express them in decreasing order, separated by commas.\"\n",
|
||||
"mathproxyagent.initiate_chat(assistant, math_problem)"
|
||||
]
|
||||
},
|
||||
{
|
||||
@ -760,11 +753,8 @@
|
||||
"assistant.reset() # clear LLM assistant's message history\n",
|
||||
"\n",
|
||||
"# we set the prompt_type to \"python\", which is a simplied version of the default prompt.\n",
|
||||
"math_problem = \"Problem: If $725x + 727y = 1500$ and $729x+ 731y = 1508$, what is the value of $x - y$ ?\"\n",
|
||||
"assistant.receive(\n",
|
||||
" mathproxyagent.generate_init_prompt(math_problem, prompt_type=\"python\"),\n",
|
||||
" mathproxyagent,\n",
|
||||
")"
|
||||
"math_problem = \"Problem: If $725x + 727y = 1500$ and $729x+ 731y = 1508$, what is the value of $x - y$ ?\"\n",
|
||||
"mathproxyagent.initiate_chat(assistant, math_problem, prompt_type=\"python\")"
|
||||
]
|
||||
},
|
||||
{
|
||||
@ -904,11 +894,8 @@
|
||||
" os.environ[\"WOLFRAM_ALPHA_APPID\"] = open(\"wolfram.txt\").read().strip()\n",
|
||||
"\n",
|
||||
"# we set the prompt_type to \"two_tools\", which allows the assistant to select wolfram alpha when necessary.\n",
|
||||
"math_problem = \"Find all numbers $a$ for which the graph of $y=x^2+a$ and the graph of $y=ax$ intersect. Express your answer in interval notation.\"\n",
|
||||
"assistant.receive(\n",
|
||||
" mathproxyagent.generate_init_prompt(math_problem, prompt_type=\"two_tools\"),\n",
|
||||
" mathproxyagent,\n",
|
||||
")"
|
||||
"math_problem = \"Find all numbers $a$ for which the graph of $y=x^2+a$ and the graph of $y=ax$ intersect. Express your answer in interval notation.\"\n",
|
||||
"mathproxyagent.initiate_chat(assistant, math_problem, prompt_type=\"two_tools\")"
|
||||
]
|
||||
}
|
||||
],
|
||||
|
||||
@ -44,7 +44,7 @@
|
||||
},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"# %pip install flaml[autogen]==2.0.0rc3"
|
||||
"# %pip install flaml[autogen]~=2.0.0rc4"
|
||||
]
|
||||
},
|
||||
{
|
||||
@ -209,9 +209,9 @@
|
||||
" use_docker=False, # set to True if you are using docker\n",
|
||||
")\n",
|
||||
"# the assistant receives a message from the user, which contains the task description\n",
|
||||
"assistant.receive(\n",
|
||||
"user.initiate_chat(\n",
|
||||
" assistant,\n",
|
||||
" \"\"\"What date is today? Compare the year-to-date gain for META and TESLA.\"\"\",\n",
|
||||
" user,\n",
|
||||
")"
|
||||
]
|
||||
},
|
||||
@ -314,9 +314,9 @@
|
||||
],
|
||||
"source": [
|
||||
"# followup of the previous question\n",
|
||||
"assistant.receive(\n",
|
||||
" \"\"\"Plot a chart of their stock price change YTD and save to stock_price_ytd.png.\"\"\",\n",
|
||||
" user\n",
|
||||
"user.send(\n",
|
||||
" recipient=assistant,\n",
|
||||
" message=\"\"\"Plot a chart of their stock price change YTD and save to stock_price_ytd.png.\"\"\",\n",
|
||||
")"
|
||||
]
|
||||
},
|
||||
|
||||
@ -36,7 +36,7 @@
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"# %pip install flaml[mathchat]==2.0.0rc3"
|
||||
"# %pip install flaml[mathchat]~=2.0.0rc4"
|
||||
]
|
||||
},
|
||||
{
|
||||
@ -79,11 +79,11 @@
|
||||
"source": [
|
||||
"## Making Function Calls\n",
|
||||
"\n",
|
||||
"In this example, we demonstrate function call execution with `AssistantAgent` and `UserProxyAgent`. With the default system prompt of `AssistantAgent`, we allow the LLM assistant to perform tasks with code, and the `UserProxyAgent` would extract code blocks from the LLM response and execute them. With the new \"function_call\" feature, we define a new function using the pre-defined `_execute_code` from `UserProxyAgent` and specify the description of the function in the OpenAI config. \n",
|
||||
"In this example, we demonstrate function call execution with `AssistantAgent` and `UserProxyAgent`. With the default system prompt of `AssistantAgent`, we allow the LLM assistant to perform tasks with code, and the `UserProxyAgent` would extract code blocks from the LLM response and execute them. With the new \"function_call\" feature, we define a new function using the pre-defined `execute_code` from `UserProxyAgent` and specify the description of the function in the OpenAI config. \n",
|
||||
"\n",
|
||||
"Then, the model has two paths to execute code:\n",
|
||||
"1. Put the code blocks in the response. `UserProxyAgent` will extract and execute the code through `_execute_code` method in the class.\n",
|
||||
"2. As we put a function description to OpenAI config and passed a function `execute_code_function` to `UserProxyAgent`, the model can also make function calls (will be put in `function_call` field of the API reply). `UserProxyAgent` will execute the function call through a `_execute_function` method."
|
||||
"1. Put the code blocks in the response. `UserProxyAgent` will extract and execute the code through `execute_code` method in the class.\n",
|
||||
"2. As we put a function description to OpenAI config and register a function `exec_code` in `UserProxyAgent`, the model can also make function calls (will be put in `function_call` field of the API reply). `UserProxyAgent` will execute the function call through the registered `exec_code` method."
|
||||
]
|
||||
},
|
||||
{
|
||||
@ -234,24 +234,26 @@
|
||||
"}\n",
|
||||
"chatbot = AssistantAgent(\"chatbot\", config_list=config_list, **oai_config)\n",
|
||||
"\n",
|
||||
"# use pre-defined execute_code function from a UserProxyAgent instance\n",
|
||||
"# for simplicity, we don't pass in `exec_func` directly to UserProxyAgent because it requires a list of tuple as parameter\n",
|
||||
"# instead, we define a wrapper function to call `exec_func`\n",
|
||||
"exec_func = UserProxyAgent(name=\"execute_code\", work_dir=\"coding\", use_docker=False)._execute_code\n",
|
||||
"\n",
|
||||
"def execute_code(code_type, code):\n",
|
||||
" return exec_func([(code_type, code)])\n",
|
||||
"\n",
|
||||
"# create a UserProxyAgent instance named \"user\"\n",
|
||||
"user = UserProxyAgent(\n",
|
||||
" \"user\",\n",
|
||||
" human_input_mode=\"NEVER\",\n",
|
||||
" function_map={\"execute_code\": execute_code},\n",
|
||||
" work_dir=\"coding\",\n",
|
||||
")\n",
|
||||
"\n",
|
||||
"# define an `execute_code` function according to the function desription\n",
|
||||
"def exec_code(code_type, code):\n",
|
||||
" # here we reuse the method in the user proxy agent\n",
|
||||
" # in general, this is not necessary\n",
|
||||
" return user.execute_code([(code_type, code)])\n",
|
||||
"\n",
|
||||
"# register the `execute_code` function\n",
|
||||
"user.register_function(function_map={\"execute_code\": exec_code})\n",
|
||||
"\n",
|
||||
"# start the conversation\n",
|
||||
"chatbot.receive(\n",
|
||||
"user.initiate_chat(\n",
|
||||
" chatbot,\n",
|
||||
" \"Draw a rocket and save to a file named 'rocket.svg'\",\n",
|
||||
" user,\n",
|
||||
")\n"
|
||||
]
|
||||
},
|
||||
@ -289,7 +291,7 @@
|
||||
"source": [
|
||||
"## Another example with Wolfram Alpha API\n",
|
||||
"\n",
|
||||
"We give another example of query Wolfram Alpha API to solve math problem. We use the predefined function from `MathUserProxyAgent()`, we directly pass the class method, `MathUserProxyAgent()._execute_one_wolfram_query`, as the function to be called."
|
||||
"We give another example of query Wolfram Alpha API to solve math problem. We use the predefined function `MathUserProxyAgent().execute_one_wolfram_query` as the function to be called."
|
||||
]
|
||||
},
|
||||
{
|
||||
@ -389,19 +391,18 @@
|
||||
"}\n",
|
||||
"chatbot = AssistantAgent(\"chatbot\", sys_prompt, config_list=config_list, **oai_config)\n",
|
||||
"\n",
|
||||
"\n",
|
||||
"# the key in `function_map` should match the function name passed to OpenAI\n",
|
||||
"# we pass a class instance directly\n",
|
||||
"# the key in `function_map` should match the function name in \"functions\" above\n",
|
||||
"# we register a class instance method directly\n",
|
||||
"user = UserProxyAgent(\n",
|
||||
" \"user\",\n",
|
||||
" human_input_mode=\"NEVER\",\n",
|
||||
" function_map={\"query_wolfram\": MathUserProxyAgent()._execute_one_wolfram_query},\n",
|
||||
" function_map={\"query_wolfram\": MathUserProxyAgent().execute_one_wolfram_query},\n",
|
||||
")\n",
|
||||
"\n",
|
||||
"# start the conversation\n",
|
||||
"chatbot.receive(\n",
|
||||
"user.initiate_chat(\n",
|
||||
" chatbot,\n",
|
||||
" \"Problem: Find all $x$ that satisfy the inequality $(2x+10)(x+3)<(3x+9)(x+8)$. Express your answer in interval notation.\",\n",
|
||||
" user,\n",
|
||||
")\n"
|
||||
]
|
||||
}
|
||||
|
||||
@ -44,7 +44,7 @@
|
||||
},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"# %pip install flaml[autogen]==2.0.0rc3"
|
||||
"# %pip install flaml[autogen]~=2.0.0rc4"
|
||||
]
|
||||
},
|
||||
{
|
||||
|
||||
@ -44,7 +44,7 @@
|
||||
},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"# %pip install flaml[autogen]==2.0.0rc3"
|
||||
"# %pip install flaml[autogen]~=2.0.0rc4"
|
||||
]
|
||||
},
|
||||
{
|
||||
@ -419,9 +419,9 @@
|
||||
],
|
||||
"source": [
|
||||
"# the assistant receives a message from the user, which contains the task description\n",
|
||||
"assistant.receive(\n",
|
||||
"user.initiate_chat(\n",
|
||||
" assistant,\n",
|
||||
" \"\"\"Suggest a fix to an open good first issue of flaml\"\"\",\n",
|
||||
" user\n",
|
||||
")"
|
||||
]
|
||||
},
|
||||
|
||||
@ -48,7 +48,7 @@
|
||||
},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"# %pip install flaml[autogen]==2.0.0rc3"
|
||||
"# %pip install flaml[autogen]~=2.0.0rc4"
|
||||
]
|
||||
},
|
||||
{
|
||||
|
||||
@ -82,7 +82,7 @@
|
||||
"source": [
|
||||
"from flaml import oai\n",
|
||||
"\n",
|
||||
"config_list = oai.config_list_openai_aoai(exclude=\"openai\")"
|
||||
"config_list = oai.config_list_openai_aoai()"
|
||||
]
|
||||
},
|
||||
{
|
||||
|
||||
@ -311,7 +311,7 @@ def test_math(num_samples=-1):
|
||||
% data["problem"]
|
||||
]
|
||||
|
||||
oai.ChatCompletion.set_cache(seed)
|
||||
oai.Completion.set_cache(seed)
|
||||
vanilla_config = {
|
||||
"model": "text-davinci-003",
|
||||
"temperature": 1,
|
||||
@ -321,8 +321,8 @@ def test_math(num_samples=-1):
|
||||
"stop": "###",
|
||||
}
|
||||
test_data_sample = test_data[0:3]
|
||||
result = oai.ChatCompletion.test(test_data_sample, eval_math_responses, **vanilla_config)
|
||||
result = oai.ChatCompletion.test(
|
||||
result = oai.Completion.test(test_data_sample, eval_math_responses, **vanilla_config)
|
||||
result = oai.Completion.test(
|
||||
test_data_sample,
|
||||
eval_math_responses,
|
||||
agg_method="median",
|
||||
@ -335,13 +335,13 @@ def test_math(num_samples=-1):
|
||||
def my_average(results):
|
||||
return np.mean(results)
|
||||
|
||||
result = oai.ChatCompletion.test(
|
||||
result = oai.Completion.test(
|
||||
test_data_sample,
|
||||
eval_math_responses,
|
||||
agg_method=my_median,
|
||||
**vanilla_config,
|
||||
)
|
||||
result = oai.ChatCompletion.test(
|
||||
result = oai.Completion.test(
|
||||
test_data_sample,
|
||||
eval_math_responses,
|
||||
agg_method={
|
||||
@ -355,7 +355,7 @@ def test_math(num_samples=-1):
|
||||
|
||||
print(result)
|
||||
|
||||
config, _ = oai.ChatCompletion.tune(
|
||||
config, _ = oai.Completion.tune(
|
||||
data=tune_data, # the data for tuning
|
||||
metric="expected_success", # the metric to optimize
|
||||
mode="max", # the optimization mode
|
||||
@ -368,7 +368,7 @@ def test_math(num_samples=-1):
|
||||
stop="###", # the stop sequence
|
||||
)
|
||||
print("tuned config", config)
|
||||
result = oai.ChatCompletion.test(test_data_sample, config_list=oai.config_list_openai_aoai(KEY_LOC), **config)
|
||||
result = oai.Completion.test(test_data_sample, config_list=oai.config_list_openai_aoai(KEY_LOC), **config)
|
||||
print("result from tuned config:", result)
|
||||
print("empty responses", eval_math_responses([], None))
|
||||
|
||||
|
||||
@ -19,8 +19,8 @@ def test_agent():
|
||||
dummy_agent_1._oai_conversations["dummy_agent_2"]
|
||||
), "When the message is not an valid openai message, it should not be appended to the oai conversation."
|
||||
|
||||
dummy_agent_1._send("hello", dummy_agent_2) # send a str
|
||||
dummy_agent_1._send(
|
||||
dummy_agent_1.send("hello", dummy_agent_2) # send a str
|
||||
dummy_agent_1.send(
|
||||
{
|
||||
"content": "hello",
|
||||
},
|
||||
@ -29,7 +29,7 @@ def test_agent():
|
||||
|
||||
# receive dict with no openai fields
|
||||
pre_len = len(dummy_agent_1._oai_conversations["dummy_agent_2"])
|
||||
dummy_agent_1._send({"message": "hello"}, dummy_agent_2) # send dict with wrong field
|
||||
dummy_agent_1.send({"message": "hello"}, dummy_agent_2) # send dict with wrong field
|
||||
|
||||
assert pre_len == len(
|
||||
dummy_agent_1._oai_conversations["dummy_agent_2"]
|
||||
|
||||
@ -29,12 +29,12 @@ def test_gpt35(human_input_mode="NEVER", max_consecutive_auto_reply=5):
|
||||
timeout=60,
|
||||
)
|
||||
coding_task = "Print hello world to a file called hello.txt"
|
||||
assistant.receive(coding_task, user)
|
||||
user.initiate_chat(assistant, coding_task)
|
||||
# coding_task = "Create a powerpoint with the text hello world in it."
|
||||
# assistant.receive(coding_task, user)
|
||||
assistant.reset()
|
||||
coding_task = "Save a pandas df with 3 rows and 3 columns to disk."
|
||||
assistant.receive(coding_task, user)
|
||||
user.initiate_chat(assistant, coding_task)
|
||||
assert not isinstance(user.use_docker, bool) # None or str
|
||||
|
||||
|
||||
@ -54,21 +54,21 @@ def test_create_execute_script(human_input_mode="NEVER", max_consecutive_auto_re
|
||||
max_consecutive_auto_reply=max_consecutive_auto_reply,
|
||||
is_termination_msg=lambda x: x.get("content", "").rstrip().endswith("TERMINATE"),
|
||||
)
|
||||
assistant.receive(
|
||||
user.initiate_chat(
|
||||
assistant,
|
||||
"""Create and execute a script to plot a rocket without using matplotlib""",
|
||||
user,
|
||||
)
|
||||
assistant.reset()
|
||||
assistant.receive(
|
||||
user.initiate_chat(
|
||||
assistant,
|
||||
"""Create a temp.py file with the following content:
|
||||
```
|
||||
print('Hello world!')
|
||||
```""",
|
||||
user,
|
||||
)
|
||||
print(conversations)
|
||||
oai.ChatCompletion.start_logging(compact=False)
|
||||
assistant.receive("""Execute temp.py""", user)
|
||||
user.send("""Execute temp.py""", assistant)
|
||||
print(oai.ChatCompletion.logged_history)
|
||||
oai.ChatCompletion.stop_logging()
|
||||
|
||||
@ -86,26 +86,33 @@ def test_tsp(human_input_mode="NEVER", max_consecutive_auto_reply=10):
|
||||
"Can we add a new point to the graph? It's distance should be randomly between 0 - 5 to each of the existing points.",
|
||||
]
|
||||
|
||||
class TSPUserProxyAgent(UserProxyAgent):
|
||||
def __init__(self, *args, **kwargs):
|
||||
super().__init__(*args, **kwargs)
|
||||
with open(f"{here}/tsp_prompt.txt", "r") as f:
|
||||
self._prompt = f.read()
|
||||
|
||||
def generate_init_prompt(self, question) -> str:
|
||||
return self._prompt.format(question=question)
|
||||
|
||||
oai.ChatCompletion.start_logging()
|
||||
assistant = AssistantAgent("assistant", temperature=0, config_list=config_list)
|
||||
user = UserProxyAgent(
|
||||
user = TSPUserProxyAgent(
|
||||
"user",
|
||||
work_dir=f"{here}",
|
||||
human_input_mode=human_input_mode,
|
||||
max_consecutive_auto_reply=max_consecutive_auto_reply,
|
||||
)
|
||||
with open(f"{here}/tsp_prompt.txt", "r") as f:
|
||||
prompt = f.read()
|
||||
# agent.receive(prompt.format(question=hard_questions[0]), user)
|
||||
# agent.receive(prompt.format(question=hard_questions[1]), user)
|
||||
assistant.receive(prompt.format(question=hard_questions[2]), user)
|
||||
user.initiate_chat(assistant, question=hard_questions[2])
|
||||
print(oai.ChatCompletion.logged_history)
|
||||
oai.ChatCompletion.stop_logging()
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
test_gpt35()
|
||||
test_create_execute_script(human_input_mode="TERMINATE")
|
||||
# test_gpt35()
|
||||
# test_create_execute_script(human_input_mode="TERMINATE")
|
||||
# when GPT-4, i.e., the DEFAULT_MODEL, is used, conversation in the following test
|
||||
# should terminate in 2-3 rounds of interactions (because is_termination_msg should be true after 2-3 rounds)
|
||||
# although the max_consecutive_auto_reply is set to 10.
|
||||
|
||||
@ -64,24 +64,24 @@ def test_execute_one_python_code():
|
||||
|
||||
# no output found 1
|
||||
code = "x=3"
|
||||
assert mathproxyagent._execute_one_python_code(code)[0] == "No output found. Make sure you print the results."
|
||||
assert mathproxyagent.execute_one_python_code(code)[0] == "No output found. Make sure you print the results."
|
||||
|
||||
# no output found 2
|
||||
code = "if 4 > 5:\n\tprint('True')"
|
||||
|
||||
assert mathproxyagent._execute_one_python_code(code)[0] == "No output found."
|
||||
assert mathproxyagent.execute_one_python_code(code)[0] == "No output found."
|
||||
|
||||
# return error
|
||||
code = "2+'2'"
|
||||
assert "Error:" in mathproxyagent._execute_one_python_code(code)[0]
|
||||
assert "Error:" in mathproxyagent.execute_one_python_code(code)[0]
|
||||
|
||||
# save previous status
|
||||
mathproxyagent._execute_one_python_code("x=3\ny=x*2")
|
||||
assert mathproxyagent._execute_one_python_code("print(y)")[0].strip() == "6"
|
||||
mathproxyagent.execute_one_python_code("x=3\ny=x*2")
|
||||
assert mathproxyagent.execute_one_python_code("print(y)")[0].strip() == "6"
|
||||
|
||||
code = "print('*'*2001)"
|
||||
assert (
|
||||
mathproxyagent._execute_one_python_code(code)[0]
|
||||
mathproxyagent.execute_one_python_code(code)[0]
|
||||
== "Your requested query response is too long. You might have made a mistake. Please revise your reasoning and query."
|
||||
)
|
||||
|
||||
@ -91,7 +91,7 @@ def test_execute_one_wolfram_query():
|
||||
code = "2x=3"
|
||||
|
||||
try:
|
||||
mathproxyagent._execute_one_wolfram_query(code)[0]
|
||||
mathproxyagent.execute_one_wolfram_query(code)[0]
|
||||
except ValueError:
|
||||
print("Wolfrma API key not found. Skip test.")
|
||||
|
||||
|
||||
@ -43,9 +43,9 @@ user_proxy = UserProxyAgent(
|
||||
)
|
||||
|
||||
# the assistant receives a message from the user, which contains the task description
|
||||
assistant.receive(
|
||||
user.initiate_chat(
|
||||
assistant,
|
||||
"""What date is today? Which big tech stock has the largest year-to-date gain this year? How much is the gain?""",
|
||||
user_proxy,
|
||||
)
|
||||
```
|
||||
In the example above, we create an AssistantAgent named "assistant" to serve as the assistant and a UserProxyAgent named "user_proxy" to serve as a proxy for the human user.
|
||||
@ -76,7 +76,7 @@ oai_config = {
|
||||
"functions": [
|
||||
{
|
||||
"name": "execute_code",
|
||||
"description": "Receive a list of python code or shell script and return the execution result.",
|
||||
"description": "Receive a python code or shell script and return the execution result.",
|
||||
"parameters": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
@ -99,23 +99,26 @@ oai_config = {
|
||||
# create an AssistantAgent instance named "assistant"
|
||||
chatbot = AssistantAgent("assistant", config_list=config_list, **oai_config)
|
||||
|
||||
# define your own function. Here we use a pre-defined '_execute_code' function from a UserProxyAgent instance
|
||||
# we define a wrapper function to call `exec_func`
|
||||
exec_func = UserProxyAgent(name="execute_code", work_dir="coding", use_docker=False)._execute_code
|
||||
def execute_code(code_type, code):
|
||||
return exec_func([(code_type, code)])
|
||||
|
||||
# create a UserProxyAgent instance named "user", the execute_code_function is passed
|
||||
# create a UserProxyAgent instance named "user"
|
||||
user = UserProxyAgent(
|
||||
"user",
|
||||
human_input_mode="NEVER",
|
||||
function_map={"execute_code": execute_code},
|
||||
work_dir="coding",
|
||||
)
|
||||
|
||||
# define an `execute_code` function according to the function desription
|
||||
def execute_code(code_type, code):
|
||||
# here we reuse the method in the user proxy agent
|
||||
# in general, this is not necessary
|
||||
return user.execute_code([(code_type, code)])
|
||||
|
||||
# register the `execute_code` function
|
||||
user.register_function(function_map={"execute_code": execute_code})
|
||||
|
||||
# start the conversation
|
||||
chatbot.receive(
|
||||
user.initiate_chat(
|
||||
assistant,
|
||||
"Draw a rocket and save to a file named 'rocket.svg'",
|
||||
user,
|
||||
)
|
||||
```
|
||||
|
||||
|
||||
Loading…
x
Reference in New Issue
Block a user