Cannot import name 'version_short' from 'pydantic.version'

[liuxiaojiao]

Checked other resources

• I added a very descriptive title to this question.
• I searched the LangChain documentation with the integrated search.
• I used the GitHub search to find a similar question and didn't find it.

Commit to Help

• I commit to help with one of those options 👆

Example Code

%pip install langchain
%pip install openai
%pip install langchain-openai

from langchain_openai import OpenAI

Description

I'm trying to install langchain in AWS SageMaker Studio.

There is an error in from langchain_openai import OpenAI: ImportError: cannot import name 'version_short' from 'pydantic.version'

System Info

python version: 3.10.13

[dosubot[bot]]

🤖

Hey there @liuxiaojiao! 🎉 Good to see you back. How's it going?

The error message you're encountering suggests that there's an issue with the version of the Pydantic library installed in your AWS SageMaker Studio environment. LangChain and the OpenAI library may require a specific version of Pydantic to function correctly. If the installed version of Pydantic in your environment is not compatible with these libraries, it could lead to import errors like the one you're experiencing.

To resolve this issue, you should check the required Pydantic version for LangChain and OpenAI, and ensure that the correct version is installed in your AWS SageMaker Studio environment. You can do this by using the pip command to check the installed version of Pydantic and to install the correct version if necessary.

Here is an example of how you can check the installed version of Pydantic:

!pip show pydantic

And here is an example of how you can install a specific version of Pydantic:

!pip install pydantic==<required_version>

Replace <required_version> with the version number required by LangChain and OpenAI.

According to the pyproject.toml files in the LangChain repository, LangChain requires any version of Pydantic that is greater than or equal to 1 and less than 3. You can find this information in the following sources:

• libs/cli/langchain_cli/project_template/pyproject.toml
• libs/langchain/pyproject.toml

Please note that the actual code you need to run may vary depending on the specific requirements of LangChain and OpenAI, and the configuration of your AWS SageMaker Studio environment.

Sources

• libs/partners/openai/langchain_openai/llms/base.py
• libs/cli/langchain_cli/project_template/pyproject.toml
• libs/langchain/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.

[cxlhyx]

This is a version error. Delete the installed version and specify a new version for installation. Solution:

pip uninstall pydantic
pip install pydantic==1.10.8

[niketdoshi]

Thanks for the reply!

I had installed the latest version of LangChain using mamba install, which installs pydantic==2.6.1 by default. Downgrading it to v1.10.8 fixes the error.

[cxlhyx]

Yes, you're right. This is a version issue, so it can be resolved.

[tedsecretsource]

Downgrading may have fixed it but definitely introduced other issues along the way that made the system unusable. We suspect this is related to updates to LangChain in recent months but lack concrete evidence.

[athewsey]

Also came across this recently in SageMaker myself and tried to dig around: I think there's something weird going on with their installed pydantic version, as documented at aws/sagemaker-distribution#436

For me, explicitly re-installing either pydantic 1.10.14 or 2.7.0 (both of which are kinda-sorta installed already - one via pip, the other via conda) seemed to resolve the ImportError.