AttributeError: 'AIMessageChunk' object has no attribute 'text'

[FloTeu]

Checked other resources

• I added a very descriptive title to this issue.
• I searched the LangChain documentation with the integrated search.
• I used the GitHub search to find a similar question and didn't find it.
• I am sure that this is a bug in LangChain rather than my code.

Example Code

install the following dependencies

pip install openai
pip install google-search-results
pip install langchain # version 0.1.9
pip install numexpr

Run the following python code (and add the openai api key):

from langchain import hub
from langchain import LLMMathChain, SerpAPIWrapper
from langchain.agents import Tool
from langchain.chat_models import ChatOpenAI
from langchain.agents import AgentExecutor, create_openai_functions_agent

import os
os.environ['OPENAI_API_KEY'] = str("xxx")

llm = ChatOpenAI(temperature=0, model="gpt-3.5-turbo-0613")
llm_math_chain = LLMMathChain.from_llm(llm=llm, verbose=True)
tools = [
    Tool(
        name="Calculator",
        func=llm_math_chain.run,
        description="useful for when you need to answer questions about math"
    )
]

hub_prompt: object = hub.pull("hwchase17/openai-tools-agent")
agent = create_openai_functions_agent(llm, tools, hub_prompt)
agent_executor = AgentExecutor(agent=agent, tools=tools, verbose=True, handle_parsing_errors=True)
input_prompt = "What is the square root of the year of birth of the founder of Space X?"
agent_executor.invoke({"input": input_prompt})

Error Message and Stack Trace (if applicable)

AttributeError: 'AIMessageChunk' object has no attribute 'text'

Trace
Traceback (most recent call last):
File "/Users/fteutsch/Desktop/PythonProjects/private/digiprod-gen/bug_report.py", line 31, in
agent_executor.invoke({"input": input_prompt})
File "/Users/fteutsch/Library/Caches/pypoetry/virtualenvs/digiprod-gen-qSAcCc-Q-py3.10/lib/python3.10/site-packages/langchain/chains/base.py", line 163, in invoke
raise e
File "/Users/fteutsch/Library/Caches/pypoetry/virtualenvs/digiprod-gen-qSAcCc-Q-py3.10/lib/python3.10/site-packages/langchain/chains/base.py", line 153, in invoke
self._call(inputs, run_manager=run_manager)
File "/Users/fteutsch/Library/Caches/pypoetry/virtualenvs/digiprod-gen-qSAcCc-Q-py3.10/lib/python3.10/site-packages/langchain/agents/agent.py", line 1391, in _call
next_step_output = self._take_next_step(
File "/Users/fteutsch/Library/Caches/pypoetry/virtualenvs/digiprod-gen-qSAcCc-Q-py3.10/lib/python3.10/site-packages/langchain/agents/agent.py", line 1097, in _take_next_step
[
File "/Users/fteutsch/Library/Caches/pypoetry/virtualenvs/digiprod-gen-qSAcCc-Q-py3.10/lib/python3.10/site-packages/langchain/agents/agent.py", line 1097, in
[
File "/Users/fteutsch/Library/Caches/pypoetry/virtualenvs/digiprod-gen-qSAcCc-Q-py3.10/lib/python3.10/site-packages/langchain/agents/agent.py", line 1125, in _iter_next_step
output = self.agent.plan(
File "/Users/fteutsch/Library/Caches/pypoetry/virtualenvs/digiprod-gen-qSAcCc-Q-py3.10/lib/python3.10/site-packages/langchain/agents/agent.py", line 387, in plan
for chunk in self.runnable.stream(inputs, config={"callbacks": callbacks}):
File "/Users/fteutsch/Library/Caches/pypoetry/virtualenvs/digiprod-gen-qSAcCc-Q-py3.10/lib/python3.10/site-packages/langchain_core/runnables/base.py", line 2427, in stream
yield from self.transform(iter([input]), config, **kwargs)
File "/Users/fteutsch/Library/Caches/pypoetry/virtualenvs/digiprod-gen-qSAcCc-Q-py3.10/lib/python3.10/site-packages/langchain_core/runnables/base.py", line 2414, in transform
yield from self._transform_stream_with_config(
File "/Users/fteutsch/Library/Caches/pypoetry/virtualenvs/digiprod-gen-qSAcCc-Q-py3.10/lib/python3.10/site-packages/langchain_core/runnables/base.py", line 1494, in _transform_stream_with_config
chunk: Output = context.run(next, iterator) # type: ignore
File "/Users/fteutsch/Library/Caches/pypoetry/virtualenvs/digiprod-gen-qSAcCc-Q-py3.10/lib/python3.10/site-packages/langchain_core/runnables/base.py", line 2378, in _transform
for output in final_pipeline:
File "/Users/fteutsch/Library/Caches/pypoetry/virtualenvs/digiprod-gen-qSAcCc-Q-py3.10/lib/python3.10/site-packages/langchain_core/runnables/base.py", line 1032, in transform
for chunk in input:
File "/Users/fteutsch/Library/Caches/pypoetry/virtualenvs/digiprod-gen-qSAcCc-Q-py3.10/lib/python3.10/site-packages/langchain_core/runnables/base.py", line 4167, in transform
yield from self.bound.transform(
File "/Users/fteutsch/Library/Caches/pypoetry/virtualenvs/digiprod-gen-qSAcCc-Q-py3.10/lib/python3.10/site-packages/langchain_core/runnables/base.py", line 1042, in transform
yield from self.stream(final, config, **kwargs)
File "/Users/fteutsch/Library/Caches/pypoetry/virtualenvs/digiprod-gen-qSAcCc-Q-py3.10/lib/python3.10/site-packages/langchain_core/language_models/chat_models.py", line 250, in stream
raise e
File "/Users/fteutsch/Library/Caches/pypoetry/virtualenvs/digiprod-gen-qSAcCc-Q-py3.10/lib/python3.10/site-packages/langchain_core/language_models/chat_models.py", line 234, in stream
for chunk in self._stream(
File "/Users/fteutsch/Library/Caches/pypoetry/virtualenvs/digiprod-gen-qSAcCc-Q-py3.10/lib/python3.10/site-packages/langchain_community/chat_models/openai.py", line 418, in _stream
run_manager.on_llm_new_token(chunk.text, chunk=cg_chunk)
AttributeError: 'AIMessageChunk' object has no attribute 'text'

Description

Since i updated langchain from 0.1.7 to latest version 0.1.9 i got the exception mentioned above.
I already found the issue and maybe the solution as well.

libs/community/langchain_community/chat_models/openai.py
lines 414-418

cg_chunk = ChatGenerationChunk(
    message=chunk, generation_info=generation_info
)
if run_manager:
    run_manager.on_llm_new_token(**chunk**.text, chunk=cg_chunk)

chunk has the type AIMessageChunk which does not contain the attribute text, whereas cg_chunk has the type ChatGenerationChunk which contains text as attribute (and in version 0.1.7 the same class was used).
The fix probably would be:

cg_chunk = ChatGenerationChunk(
    message=chunk, generation_info=generation_info
)
if run_manager:
    run_manager.on_llm_new_token(**cg_chunk**.text, chunk=cg_chunk)

line 510 in the same file contains the same issue.

System Info

langchain==0.1.9
langchain-community==0.0.22
langchain-core==0.1.26
langchainhub==0.1.14

macOs
python 3.10

[eyurtsev]

confirmed

[eyurtsev]

Fixed by https://github.com/langchain-ai/langchain/pull/18006/files

[eyurtsev]

Fix merged. Closing issue -- will be available in next release.

@FloTeu we recommend transitioning into the partner package for open ai rather than using the community version.

[magedhelmy1]

Fix merged. Closing issue -- will be available in next release.

@FloTeu we recommend transitioning into the partner package for open ai rather than using the community version.

@eyurtsev
I am also facing this issue, can you please explain what did you mean with your comment when you said partner package for open ai rather than using the community version ? it was not clear.

This is the code causing the issue:

from langchain.adapters import openai as lc_openai

async def stream_response(model, messages, temperature, max_tokens, llm_provider, websocket=None):
    paragraph = ""
    response = ""

    for chunk in lc_openai.ChatCompletion.create(
            model=model,
            messages=messages,
            temperature=temperature,
            max_tokens=max_tokens,
            provider=llm_provider,
            stream=True,
    ):
        content = chunk["choices"][0].get("delta", {}).get("content")
        if content is not None:
            response += content
            paragraph += content
            if "\n" in paragraph:
                if websocket is not None:
                    await websocket.send_json({"type": "report", "output": paragraph})
                else:
                    print(f"{Fore.GREEN}{paragraph}{Style.RESET_ALL}")
                paragraph = ""
    return response