ChromaDB Vectorstore: Customize distance calculations

[rkeshwani]

Right now the langchain chroma vectorstore doesn't allow you to adjust the metadata attribute on the create collection method of the ChromaDB client so you can't adjust the formula for distance calculations.

Chroma DB introduced the ability to add metadata to collections to tell the index which distance calculation is used in release https://github.com/chroma-core/chroma/releases/tag/0.3.15

Specifically in this pull request: chroma-core/chroma#245

Langchain doesn't provide a way to adjust this vectorstore's distance calculation formula.
Referenced here: https://github.com/hwchase17/langchain/blob/2f49c96532725fdb48ea11417270245e694574d1/langchain/vectorstores/chroma.py#L84

[rkeshwani]

Created Pull request to add this ability. Tests appeared to pass but I didn't see a test specific for ChromaDB to fully test this functionality.

[jasonslyvia]

ping, would love to use different distance methods other than the defaulted 'l2'

[dosubot]

Hi, @rkeshwani! I'm Dosu, and I'm here to help the LangChain team manage their backlog. I wanted to let you know that we are marking this issue as stale.

From what I understand, the issue is about the inability to customize distance calculations in the ChromaDB Vectorstore of the Langchain project. You have created a pull request to add this ability, but there is a need for a specific test for ChromaDB to fully test this functionality. Another user has also expressed interest in using different distance methods and has received positive reactions from other users.

Before we close this issue, we wanted to check with you if it is still relevant to the latest version of the LangChain repository. If it is, please let us know by commenting on the issue. Otherwise, feel free to close the issue yourself or it will be automatically closed in 7 days.

Thank you for your contribution to the LangChain project!