InvalidRequestError: Invalid URL (POST /v1/openai/deployments/gpt-35-turbo/chat/completions)

[AugustusLZJ]

I am seeing the following in issues in several notebooks, ./notebook/agentchat_MathChat.ipynb is one of them.

I am setting:

config_list = [
    {
        'model': 'gpt-3.5-turbo',
        'api_key': token_response.json()["access_token"],
        'base_url': 'https://chat-my.linkxyz.com/',
        'api_type': "azure",
        'api_version': "2023-08-01-preview"
    },
]

It seems AssistantAgent is good, but while calling:

math_problem = "Find all $x$ that satisfy the inequality $(2x+10)(x+3)<(3x+9)(x+8)$. Express your answer in interval notation."
mathproxyagent.initiate_chat(assistant, problem=math_problem)

It shows up the following error:

Let's use Python to solve a math problem.

Query requirements:
You should always use the 'print' function for the output and use fractions/radical forms instead of decimals.
You can use packages like sympy to help you.
You must follow the formats below to write your code:

python
# your code


First state the key idea to solve the problem. You may choose from three ways to solve the problem:
Case 1: If the problem can be solved with Python code directly, please write a program to solve it. You can enumerate all possible arrangements if needed.
Case 2: If the problem is mostly reasoning, you can solve it by yourself directly.
Case 3: If the problem cannot be handled in the above two ways, please follow this process:
1. Solve the problem step by step (do not over-divide the steps).
2. Take out any queries that can be asked through Python (for example, any calculations or equations that can be calculated).
3. Wait for me to give the results.
4. Continue if you think the result is correct. If the result is invalid or unexpected, please correct your query or reasoning.

After all the queries are run and you get the answer, put the answer in \boxed{}.

Problem:
Find all $x$ that satisfy the inequality $(2x+10)(x+3)<(3x+9)(x+8)$. Express your answer in interval notation.

--------------------------------------------------------------------------------
---------------------------------------------------------------------------
InvalidRequestError                       Traceback (most recent call last)
/Path/autogen/notebook/agentchat_MathChat.ipynb Cell 11 line 5
      1 # given a math problem, we use the mathproxyagent to generate a prompt to be sent to the assistant as the initial message.
      2 # the assistant receives the message and generates a response. The response will be sent back to the mathproxyagent for processing.
      3 # The conversation continues until the termination condition is met, in MathChat, the termination condition is the detect of "\boxed{}" in the response.
      4 math_problem = "Find all $x$ that satisfy the inequality $(2x+10)(x+3)<(3x+9)(x+8)$. Express your answer in interval notation."
----> 5 mathproxyagent.initiate_chat(assistant, problem=math_problem)

File /Path/pyautogen/lib/python3.10/site-packages/autogen/agentchat/conversable_agent.py:531, in ConversableAgent.initiate_chat(self, recipient, clear_history, silent, **context)
    517 """Initiate a chat with the recipient agent.
    518 
    519 Reset the consecutive auto reply counter.
   (...)
    528         "message" needs to be provided if the `generate_init_message` method is not overridden.
    529 """
    530 self._prepare_chat(recipient, clear_history)
--> 531 self.send(self.generate_init_message(**context), recipient, silent=silent)

File /Path/pyautogen/lib/python3.10/site-packages/autogen/agentchat/conversable_agent.py:334, in ConversableAgent.send(self, message, recipient, request_reply, silent)
    332 valid = self._append_oai_message(message, "assistant", recipient)
    333 if valid:
--> 334     recipient.receive(message, self, request_reply, silent)
    335 else:
    336     raise ValueError(
    337         "Message can't be converted into a valid ChatCompletion message. Either content or function_call must be provided."
    338     )

File /Path/pyautogen/lib/python3.10/site-packages/autogen/agentchat/conversable_agent.py:462, in ConversableAgent.receive(self, message, sender, request_reply, silent)
    460 if request_reply is False or request_reply is None and self.reply_at_receive[sender] is False:
    461     return
--> 462 reply = self.generate_reply(messages=self.chat_messages[sender], sender=sender)
    463 if reply is not None:
    464     self.send(reply, sender, silent=silent)

File /Path/pyautogen/lib/python3.10/site-packages/autogen/agentchat/conversable_agent.py:781, in ConversableAgent.generate_reply(self, messages, sender, exclude)
    779     continue
    780 if self._match_trigger(reply_func_tuple["trigger"], sender):
--> 781     final, reply = reply_func(self, messages=messages, sender=sender, config=reply_func_tuple["config"])
    782     if final:
    783         return reply

File /Path/pyautogen/lib/python3.10/site-packages/autogen/agentchat/conversable_agent.py:606, in ConversableAgent.generate_oai_reply(self, messages, sender, config)
    603     messages = self._oai_messages[sender]
    605 # TODO: #1143 handle token limit exceeded error
--> 606 response = oai.ChatCompletion.create(
    607     context=messages[-1].pop("context", None), messages=self._oai_system_message + messages, **llm_config
    608 )
    609 return True, oai.ChatCompletion.extract_text_or_function_call(response)[0]

File /Path/pyautogen/lib/python3.10/site-packages/autogen/oai/completion.py:803, in Completion.create(cls, context, use_cache, config_list, filter_func, raise_on_ratelimit_or_timeout, allow_format_str_template, **config)
    801     base_config["max_retry_period"] = 0
    802 try:
--> 803     response = cls.create(
    804         context,
    805         use_cache,
    806         raise_on_ratelimit_or_timeout=i < last or raise_on_ratelimit_or_timeout,
    807         **base_config,
    808     )
    809     if response == -1:
    810         return response

File /Path/pyautogen/lib/python3.10/site-packages/autogen/oai/completion.py:834, in Completion.create(cls, context, use_cache, config_list, filter_func, raise_on_ratelimit_or_timeout, allow_format_str_template, **config)
    832 with diskcache.Cache(cls.cache_path) as cls._cache:
    833     cls.set_cache(seed)
--> 834     return cls._get_response(params, raise_on_ratelimit_or_timeout=raise_on_ratelimit_or_timeout)

File /Path/pyautogen/lib/python3.10/site-packages/autogen/oai/completion.py:224, in Completion._get_response(cls, config, raise_on_ratelimit_or_timeout, use_cache)
    222         response = openai_completion.create(**config)
    223     else:
--> 224         response = openai_completion.create(request_timeout=request_timeout, **config)
    225 except (
    226     ServiceUnavailableError,
    227     APIConnectionError,
    228 ):
    229     # transient error
    230     logger.info(f"retrying in {retry_wait_time} seconds...", exc_info=1)

File /Path/pyautogen/lib/python3.10/site-packages/openai/api_resources/chat_completion.py:25, in ChatCompletion.create(cls, *args, **kwargs)
     23 while True:
     24     try:
---> 25         return super().create(*args, **kwargs)
     26     except TryAgain as e:
     27         if timeout is not None and time.time() > start + timeout:

File /Path/pyautogen/lib/python3.10/site-packages/openai/api_resources/abstract/engine_api_resource.py:155, in EngineAPIResource.create(cls, api_key, api_base, api_type, request_id, api_version, organization, **params)
    129 @classmethod
    130 def create(
    131     cls,
   (...)
    138     **params,
    139 ):
    140     (
    141         deployment_id,
    142         engine,
   (...)
    152         api_key, api_base, api_type, api_version, organization, **params
    153     )
--> 155     response, _, api_key = requestor.request(
    156         "post",
    157         url,
    158         params=params,
    159         headers=headers,
    160         stream=stream,
    161         request_id=request_id,
    162         request_timeout=request_timeout,
    163     )
    165     if stream:
    166         # must be an iterator
    167         assert not isinstance(response, OpenAIResponse)

File /Path/pyautogen/lib/python3.10/site-packages/openai/api_requestor.py:299, in APIRequestor.request(self, method, url, params, headers, files, stream, request_id, request_timeout)
    278 def request(
    279     self,
    280     method,
   (...)
    287     request_timeout: Optional[Union[float, Tuple[float, float]]] = None,
    288 ) -> Tuple[Union[OpenAIResponse, Iterator[OpenAIResponse]], bool, str]:
    289     result = self.request_raw(
    290         method.lower(),
    291         url,
   (...)
    297         request_timeout=request_timeout,
    298     )
--> 299     resp, got_stream = self._interpret_response(result, stream)
    300     return resp, got_stream, self.api_key

File /Path/pyautogen/lib/python3.10/site-packages/openai/api_requestor.py:710, in APIRequestor._interpret_response(self, result, stream)
    702     return (
    703         self._interpret_response_line(
    704             line, result.status_code, result.headers, stream=True
    705         )
    706         for line in parse_stream(result.iter_lines())
    707     ), True
    708 else:
    709     return (
--> 710         self._interpret_response_line(
    711             result.content.decode("utf-8"),
    712             result.status_code,
    713             result.headers,
    714             stream=False,
    715         ),
    716         False,
    717     )

File /Path/pyautogen/lib/python3.10/site-packages/openai/api_requestor.py:775, in APIRequestor._interpret_response_line(self, rbody, rcode, rheaders, stream)
    773 stream_error = stream and "error" in resp.data
    774 if stream_error or not 200 <= rcode < 300:
--> 775     raise self.handle_error_response(
    776         rbody, rcode, resp.data, rheaders, stream_error=stream_error
    777     )
    778 return resp

InvalidRequestError: Invalid URL (POST /v1/openai/deployments/gpt-35-turbo/chat/completions)
mathproxyagent (to assistant):

Let's use Python to solve a math problem.

Query requirements:
You should always use the 'print' function for the output and use fractions/radical forms instead of decimals.
You can use packages like sympy to help you.
You must follow the formats below to write your code:
```python
# your code

First state the key idea to solve the problem. You may choose from three ways to solve the problem:
Case 1: If the problem can be solved with Python code directly, please write a program to solve it. You can enumerate all possible arrangements if needed.
Case 2: If the problem is mostly reasoning, you can solve it by yourself directly.
Case 3: If the problem cannot be handled in the above two ways, please follow this process:

1. Solve the problem step by step (do not over-divide the steps).
2. Take out any queries that can be asked through Python (for example, any calculations or equations that can be calculated).
3. Wait for me to give the results.
4. Continue if you think the result is correct. If the result is invalid or unexpected, please correct your query or reasoning.

After all the queries are run and you get the answer, put the answer in \boxed{}.

Problem:
For what negative value of $k$ is there exactly one solution to the system of equations \begin{align*}
...
y &= -x + 4?
\end{align*}

---

Output is truncated. View as a scrollable element or open in a text editor. Adjust cell output settings...

InvalidRequestError Traceback (most recent call last)
/Path/autogen/notebook/agentchat_MathChat.ipynb Cell 13 line 2
1 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*}"
----> 2 mathproxyagent.initiate_chat(assistant, problem=math_problem)

File pyautogen/lib/python3.10/site-packages/autogen/agentchat/conversable_agent.py:531, in ConversableAgent.initiate_chat(self, recipient, clear_history, silent, **context)
517 """Initiate a chat with the recipient agent.
518
519 Reset the consecutive auto reply counter.
(...)
528 "message" needs to be provided if the generate_init_message method is not overridden.
529 """
530 self._prepare_chat(recipient, clear_history)
--> 531 self.send(self.generate_init_message(**context), recipient, silent=silent)

File pyautogen/lib/python3.10/site-packages/autogen/agentchat/conversable_agent.py:334, in ConversableAgent.send(self, message, recipient, request_reply, silent)
332 valid = self._append_oai_message(message, "assistant", recipient)
333 if valid:
--> 334 recipient.receive(message, self, request_reply, silent)
335 else:
336 raise ValueError(
337 "Message can't be converted into a valid ChatCompletion message. Either content or function_call must be provided."
338 )
...
776 rbody, rcode, resp.data, rheaders, stream_error=stream_error
777 )
778 return resp

InvalidRequestError: Invalid URL (POST /v1/openai/deployments/gpt-35-turbo/chat/completions)
Output is truncated. View as a scrollable element or open in a text editor. Adjust cell output settings...

[GeorgSatyros]

Seems that this issue has been addressed in the newest release v0.2.0b1.

pip install pyautogen==v0.2.0b1 should fix your issue!

[afourney]

Yes, I think specifically this PR: #551

@AugustusLZJ let us know if this addresses the issue for you, and we will close this issue.

[AugustusLZJ]

@GeorgSatyros @afourney Thank you for looking at this issue!

The v0.2.0b1 is still failing with the same error. After I installed the v0.2.0b2 I don't see the "Invalid URL" error, but I see a rate limit error instead, If the new fixes doesn't introduce rate limit errors, I guess the original "Invalid URL" error is probably resolved.

The rate limit error is like below, considering the base_url's rate limit is roughtly 6 requests per minute, is it expected to see following errors?

RateLimitError                            Traceback (most recent call last)
/Path/autogen/notebook/agentchat_auto_feedback_from_code_execution.ipynb Cell 12 line 2
      1 # followup of the previous question
----> 2 user_proxy.send(
      3     recipient=assistant,
      4     message="""Plot a chart of their stock price change YTD and save to stock_price_ytd.png.""",
      5 )

File /envs/pyautogen/lib/python3.10/site-packages/autogen/agentchat/conversable_agent.py:340, in ConversableAgent.send(self, message, recipient, request_reply, silent)
    338 valid = self._append_oai_message(message, "assistant", recipient)
    339 if valid:
--> 340     recipient.receive(message, self, request_reply, silent)
    341 else:
    342     raise ValueError(
    343         "Message can't be converted into a valid ChatCompletion message. Either content or function_call must be provided."
    344     )

File /envs/pyautogen/lib/python3.10/site-packages/autogen/agentchat/conversable_agent.py:473, in ConversableAgent.receive(self, message, sender, request_reply, silent)
    471 reply = self.generate_reply(messages=self.chat_messages[sender], sender=sender)
    472 if reply is not None:
--> 473     self.send(reply, sender, silent=silent)

File /envs/pyautogen/lib/python3.10/site-packages/autogen/agentchat/conversable_agent.py:340, in ConversableAgent.send(self, message, recipient, request_reply, silent)
    338 valid = self._append_oai_message(message, "assistant", recipient)
...
--> 926     raise self._make_status_error_from_response(err.response) from None
    927 except httpx.TimeoutException as err:
    928     if retries > 0:

RateLimitError: Error code: 429 - {'fault': {'faultstring': 'Spike arrest violation. Allowed rate : MessageRate{messagesPerPeriod=7, periodInMicroseconds=60000000, maxBurstMessageCount=1.0}', 'detail': {'errorcode': 'policies.ratelimit.SpikeArrestViolation'}}}

[AugustusLZJ]

After I adjusted the timeout in config_list, the rate limit issue is gone.

Without the noise issues, I still don't see the InvalidRequestError, which should be resolved after v0.2.0b2 being installed.

[BingkangShi]

I just faced the same problem, and i am sure it can be solved in v0.2.0b1

I use Azure OpenAI api, and the error came from the output of the teachable_agent module in Win11 terminal:

teachableagent (to user):

Hi, I'm a teachable user assistant! What's on your mind?

--------------------------------------------------------------------------------
Provide feedback to teachableagent. Press enter to skip and use auto-reply, or type 'exit' to end the conversation: hello?
user (to teachableagent):

hello?

--------------------------------------------------------------------------------
Traceback (most recent call last):
  File "...\lib\site-packages\openai\api_requestor.py", line 763, in _interpret_response_line
    raise self.handle_error_response(
openai.error.InvalidRequestError: Invalid URL (POST /v1/openai/deployments/gpt-35-turbo/chat/completions)

Following is the output of the teachable_agent module in Win11 terminal under v0.2.0b1:

Hi, I'm a teachable user assistant! What's on your mind?

--------------------------------------------------------------------------------
Provide feedback to teachableagent. Press enter to skip and use auto-reply, or type 'exit' to end the conversation: hello?
user (to teachableagent):

hello?

--------------------------------------------------------------------------------
teachableagent (to user):

Hello! How can I assist you today? Do you have any questions or tasks you need help with?

--------------------------------------------------------------------------------
Provide feedback to teachableagent. Press enter to skip and use auto-reply, or type 'exit' to end the conversation: 

thx @GeorgSatyros

[jacklatrobe]

Could be wrong, but I think it's supposed to be /deployments/modelname/completions as per https://learn.microsoft.com/en-us/azure/ai-services/openai/reference

Error shows autogen trying to hit /deployments/modelname/chat/completions

the /chat/ doesn't seem needed any more? Seems like an issue in that version.

EDIT: pip install pyautogen==v0.2.0b2 fixed this for me.

[adam-behrendt]

Hi Everyone,

I had the exact same issue, and used pip install pyautogen==v0.2.0b2.

However, the next issue that arose was: ImportError: please install flaml[openai] option to use the flaml.autogen.oai subpackage.

So of course I do: pip install 'flaml[openai]'

Which errors out with:

Requirement already satisfied: flaml[openai] in /Users/behreab1/miniconda3/lib/python3.10/site-packages (2.1.1)
Requirement already satisfied: NumPy>=1.17.0rc1 in /Users/behreab1/miniconda3/lib/python3.10/site-packages (from flaml[openai]) (1.24.3)
Collecting openai==0.27.8
  Using cached openai-0.27.8-py3-none-any.whl (73 kB)
Requirement already satisfied: diskcache in /Users/behreab1/miniconda3/lib/python3.10/site-packages (from flaml[openai]) (5.6.3)
Requirement already satisfied: tqdm in /Users/behreab1/miniconda3/lib/python3.10/site-packages (from openai==0.27.8->flaml[openai]) (4.65.0)
Requirement already satisfied: aiohttp in /Users/behreab1/miniconda3/lib/python3.10/site-packages (from openai==0.27.8->flaml[openai]) (3.9.1)
Requirement already satisfied: requests>=2.20 in /Users/behreab1/miniconda3/lib/python3.10/site-packages (from openai==0.27.8->flaml[openai]) (2.28.1)
Requirement already satisfied: charset-normalizer<3,>=2 in /Users/behreab1/miniconda3/lib/python3.10/site-packages (from requests>=2.20->openai==0.27.8->flaml[openai]) (2.0.4)
Requirement already satisfied: urllib3<1.27,>=1.21.1 in /Users/behreab1/miniconda3/lib/python3.10/site-packages (from requests>=2.20->openai==0.27.8->flaml[openai]) (1.26.15)
Requirement already satisfied: idna<4,>=2.5 in /Users/behreab1/miniconda3/lib/python3.10/site-packages (from requests>=2.20->openai==0.27.8->flaml[openai]) (3.4)
Requirement already satisfied: certifi>=2017.4.17 in /Users/behreab1/miniconda3/lib/python3.10/site-packages (from requests>=2.20->openai==0.27.8->flaml[openai]) (2023.7.22)
Requirement already satisfied: multidict<7.0,>=4.5 in /Users/behreab1/miniconda3/lib/python3.10/site-packages (from aiohttp->openai==0.27.8->flaml[openai]) (6.0.4)
Requirement already satisfied: aiosignal>=1.1.2 in /Users/behreab1/miniconda3/lib/python3.10/site-packages (from aiohttp->openai==0.27.8->flaml[openai]) (1.3.1)
Requirement already satisfied: async-timeout<5.0,>=4.0 in /Users/behreab1/miniconda3/lib/python3.10/site-packages (from aiohttp->openai==0.27.8->flaml[openai]) (4.0.3)
Requirement already satisfied: yarl<2.0,>=1.0 in /Users/behreab1/miniconda3/lib/python3.10/site-packages (from aiohttp->openai==0.27.8->flaml[openai]) (1.9.4)
Requirement already satisfied: frozenlist>=1.1.1 in /Users/behreab1/miniconda3/lib/python3.10/site-packages (from aiohttp->openai==0.27.8->flaml[openai]) (1.4.1)
Requirement already satisfied: attrs>=17.3.0 in /Users/behreab1/miniconda3/lib/python3.10/site-packages (from aiohttp->openai==0.27.8->flaml[openai]) (21.4.0)
Installing collected packages: openai
  Attempting uninstall: openai
    Found existing installation: openai 1.0.0b3
    Uninstalling openai-1.0.0b3:
      Successfully uninstalled openai-1.0.0b3
ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
pyautogen 0.2.0b2 requires openai==1.0.0b3, but you have openai 0.27.8 which is incompatible.
Successfully installed openai-0.27.8
Note: you may need to restart the kernel to use updated packages.

This is very confusing to me, as I installed openai 1.0.0b3 to work specifically with the new autogen, but it says I don't have it and downloads an older version.

I am stuck in the cycle of downloading flaml[openai] it changing my openai to an old version which is not compatible with pyautogen, to redownloading openai only to be prompted to download flaml.

I am new to this so any help would be appreciated.

Thank you.

[thinkall]

Closing this issue due to inactivity. If you have further questions, please open a new issue or join the discussion in AutoGen Discord server: https://discord.com/invite/Yb5gwGVkE5