Self query retriever with Vector Store type Elasticsearch not supported.

[hussnainahmed]

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

from langchain.schema import Document
from langchain_elasticsearch import ElasticsearchStore
from langchain.retrievers.self_query.base import SelfQueryRetriever
from langchain.chains.query_constructor.base import AttributeInfo
from langchain_community.llms import Ollama

llm = Ollama(model="llama2",base_url='url')
retriever = SelfQueryRetriever.from_llm(
    llm, vectorstore, document_content_description, metadata_field_info, verbose=True
)

Description

I am trying a retrieval query on Elasticsearch as a vector store. Embeddings and metadata are already stored in an elastic index, but I am facing an Error.

ValueError: Self query retriever with Vector Store type <class 'langchain_elasticsearch.vectorstores.ElasticsearchStore'> not supported.

System Info

$pip freeze|grep langchain
langchain==0.1.11
langchain-community==0.0.27
langchain-core==0.1.30
langchain-elasticsearch==0.1.0
langchain-text-splitters==0.0.1

[dosubot[bot]]

Hey @hussnainahmed, great to see you diving into the depths of LangChain again! Hope you're doing fantastic.

Based on the information you've provided and the context from the LangChain repository, it seems like you're encountering a ValueError because the ElasticsearchStore is not properly supported or integrated within the _get_builtin_translator function in the SelfQueryRetriever class. This function is responsible for mapping vector store types to their corresponding translator classes.

The _get_builtin_translator function in the SelfQueryRetriever class is implemented as follows:

def _get_builtin_translator(vectorstore: VectorStore) -> Visitor:
    """Get the translator class corresponding to the vector store class."""
    BUILTIN_TRANSLATORS: Dict[Type[VectorStore], Type[Visitor]] = {
        AstraDB: AstraDBTranslator,
        PGVector: PGVectorTranslator,
        Pinecone: PineconeTranslator,
        Chroma: ChromaTranslator,
        DashVector: DashvectorTranslator,
        Dingo: DingoDBTranslator,
        Weaviate: WeaviateTranslator,
        Vectara: VectaraTranslator,
        Qdrant: QdrantTranslator,
        MyScale: MyScaleTranslator,
        DeepLake: DeepLakeTranslator,
        ElasticsearchStore: ElasticsearchTranslator,
        Milvus: MilvusTranslator,
        SupabaseVectorStore: SupabaseVectorTranslator,
        TimescaleVector: TimescaleVectorTranslator,
        OpenSearchVectorSearch: OpenSearchTranslator,
        MongoDBAtlasVectorSearch: MongoDBAtlasTranslator,
    }

    if isinstance(vectorstore, Qdrant):
        return QdrantTranslator(metadata_key=vectorstore.metadata_payload_key)
    elif isinstance(vectorstore, MyScale):
        return MyScaleTranslator(metadata_key=vectorstore.metadata_column)
    elif isinstance(vectorstore, Redis):
        return RedisTranslator.from_vectorstore(vectorstore)
    elif vectorstore.__class__ in BUILTIN_TRANSLATORS:
        return BUILTIN_TRANSLATORS[vectorstore.__class__]()
    else:
        try:
            from langchain_astradb.vectorstores import AstraDBVectorStore

            if isinstance(vectorstore, AstraDBVectorStore):
                return AstraDBTranslator()
        except ImportError:
            pass

        raise ValueError(
            f"Self query retriever with Vector Store type {vectorstore.__class__}"
            f" not supported."
        )

(Source: libs/langchain/langchain/retrievers/self_query/base.py)

From the code, we can see that ElasticsearchStore is indeed mapped to ElasticsearchTranslator in the BUILTIN_TRANSLATORS dictionary. This suggests that ElasticsearchStore should be supported. The error might be triggered due to a different issue not directly visible in the provided context, such as:

1. An issue with the instantiation or configuration of the ElasticsearchStore or ElasticsearchTranslator that prevents the correct recognition or functioning of the translator.
2. A problem elsewhere in the code where the vectorstore instance's class is checked or used, leading to a mismatch or incorrect handling specific to ElasticsearchStore.

Given the provided context, the mapping between ElasticsearchStore and ElasticsearchTranslator is correctly established, suggesting that the error might stem from a different part of the system or due to a misconfiguration. Further investigation into how ElasticsearchStore instances are created, configured, and used throughout the system would be necessary to pinpoint the exact cause of the error.

I hope this helps! If you have any more questions or need further clarification, feel free to ask.

Sources

• Self query retriever with Vector Store type <class 'langchain_community.vectorstores.faiss.FAISS'> not supported.
• initializing and using a Supabase vector store using SupabaseVectorStore
• Why is this implementation of vectorstore not working?
• libs/langchain/langchain/retrievers/self_query/base.py

About Dosu

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.

[maximeperrindev]

@hussnainahmed
Try to change your ElastisearchStore import to :

from langchain_community.vectorstores.elasticsearch import ElasticsearchStore

[hussnainahmed]

Thank you @maximeperrindev. Fixed. Although there is a small typo in the call
from langchain_community.vectorstores.elasticsearch import ElasticsearchStore

[sheraz-n]

I had the same issue with pinecone. It turned out that if I import PineconeVectorStore as follows:
from langchain_pinecone import PineconeVectorStore
Then SelfQueryRetriever will not work
But if I use following import:
from langchain_community.vectorstores.pinecone import Pinecone as PineconeVectorStore
It works flawlessly

[Danouchka]

Do we have any idea when this will be fixed ?

[Danouchka]

cc @joemcelroy

[maxjakob]

This was fixed here: #19907 so in the next LangChain release @hussnainahmed's issue should disappear because it will be possible to use an ElasticsearchStore from langchain_elasticsearch.