I am trying to use Bedrock to call LLM model and integrate MongoDB as vector store.
I’ve succesfully created MongoDB client and all the codes related to Bedrock went well.
But when I try to ask questions to the model, it shows the error message below.

AttributeError: ‘Bedrock’ object has no attribute ’embed_query’

Here is my code and please advise what I should check further.

llm = Bedrock(
    model_id="amazon.titan-text-express-v1"
)

vector_search = MongoDBAtlasVectorSearch.from_connection_string(
    f"mongodb+srv://temp:{os.environ['mdb_password']}@{os.environ['cluster_name']}.mongodb.net/",
    namespace = "db_test.collection_test",
    embedding = llm
)

qa = RetrievalQA.from_chain_type(
    llm = llm, 
    chain_type = "stuff", 
    retriever = vector_search.as_retriever())

qa.invoke( "what is RAG?")

Thanks in advance.