Snowflake is a cloud-based data-warehousing platformthat allows you to store and query large amounts of data.

This page covers how to use theSnowflake ecosystem withinLangChain.

Embedding models​

Snowflake offers their open-weightarctic line of embedding models for freeonHugging Face. The most recent model, snowflake-arctic-embed-m-v1.5 featurematryoshka embedding which allows for effective vector truncation.You can use these models via theHuggingFaceEmbeddings connector:

pip install langchain-community sentence-transformers

from langchain_huggingfaceimport HuggingFaceEmbeddings

model= HuggingFaceEmbeddings(model_name="snowflake/arctic-embed-m-v1.5")

Document loader​

You can use theSnowflakeLoaderto load data from Snowflake:

from langchain_community.document_loadersimport SnowflakeLoader