Unable to use llamaIndex after openAI update

[NitishDixit]

Hello,

I have a strange situation here, I was using LlamaIndex's subQueryEngineTool to build a RAG anf it was working fine until an update from openAI, now when I try to do the following:

s_engine = SubQuestionQueryEngine.from_defaults(query_engine_tools=query_engine_tools, use_async=False) response = s_engine.query("Question about document")

I get the following error: "you tried to access openai.chatcompletion, but this is no longer supported in openai>=1.0.0 llamaindex"

I tried downgrading openAI version but then other areas of LlamIndex stop working, can someone plz help me with how can this be fixed?

Additionally, I tried using llama_index==0.8.62 but then it starts throwing error on the following code itself:
from llama_index.readers.download import download_loader from pathlib import Path PDFReader = download_loader("PDFReader")

here is the error: InvalidURL: Invalid URL 'https:/raw.githubusercontent.com/run-llama/llama-hub/main/llama_hub/library.json': No host supplied

Thanks

[dosubot[bot]]

🤖

I see that you're having trouble with the LlamaIndex's subQueryEngineTool after an update from OpenAI. The error message you're seeing suggests that the openai.chatcompletion method you're trying to access is no longer supported in the newer versions of OpenAI (>=1.0.0).

As per the dependencies of LlamaIndex v0.8.68, it requires OpenAI version >=1.1.0. This could be the reason why downgrading the OpenAI version is causing other areas of LlamaIndex to stop working.

For the second error, it seems like there's a typo in the URL 'https:/raw.githubusercontent.com/run-llama/llama-hub/main/llama_hub/library.json'. It should be 'https://raw.githubusercontent.com/run-llama/llama-hub/main/llama_hub/library.json'. The 'https://' protocol is missing an extra slash.

To resolve these issues, I would suggest the following steps:

1. Ensure that you're using the correct versions of all dependencies as listed in the pyproject.toml file. You can do this by creating a new virtual environment and installing the dependencies with the correct versions.

2. Update the URL in your code to 'https://raw.githubusercontent.com/run-llama/llama-hub/main/llama_hub/library.json' and try running your code again.

3. If the issue with openai.chatcompletion persists, you might need to refactor your code to use a different method that is supported in the newer versions of OpenAI. You can refer to the OpenAI documentation for more information on the changes made in the update and the methods that are currently supported.

Please try these steps and let me know if the issues persist.

Sources

• pyproject.toml

---

This response is meant to be useful and save you time. It is not meant to be a precise solution, but rather a starting point for your own research.

Help me be more useful! Please leave a 👍 if this is helpful and 👎 if it is irrelevant.

If you want to continue the conversation, start your reply with @dosu-bot.

[NitishDixit]

For the first error, it is bringing in openAI 1.2.4 via llama_Index itself but it still gives the completion error

[kujma10]

I have the same issue. @NitishDixit have you found a solution?

[kujma10]

I think llamaIndex needs to be migrated with openai/openai-python#742