Sub Question Query Engine

This LlamaPack inserts your data into Weaviate and uses the Sub-Question Query Engine for your RAG application.

CLI Usage

You can download llamapacks directly using llamaindex-cli, which comes installed with the llama-index python package:

llamaindex-cli download-llamapack WeaviateSubQuestionPack --download-dir ./weaviate_pack

You can then inspect the files at ./weaviate_pack and use them as a template for your own project.

Code Usage

You can download the pack to a the ./weaviate_pack directory:

from llama_index.core.llama_pack import download_llama_pack

# download and install dependencies
WeaviateSubQuestionPack = download_llama_pack(
    "WeaviateSubQuestionPack", "./weaviate_pack"
)

From here, you can use the pack, or inspect and modify the pack in ./weaviate_pack.

Then, you can set up the pack like so:

# setup pack arguments
from llama_index.core.vector_stores.types import MetadataInfo, VectorStoreInfo

vector_store_info = VectorStoreInfo(
    content_info="brief biography of celebrities",
    metadata_info=[
        MetadataInfo(
            name="category",
            type="str",
            description=(
                "Category of the celebrity, one of [Sports Entertainment, Business, Music]"
            ),
        ),
    ],
)

import weaviate

client = weaviate.Client()

nodes = [...]

# create the pack
weaviate_pack = WeaviateSubQuestion(
    collection_name="test",
    vector_store_info=vector_store_index,
    nodes=nodes,
    client=client,
)

The run() function is a light wrapper around query_engine.query().

response = weaviate_pack.run("Tell me a bout a Music celebritiy.")

You can also use modules individually.

# use the retriever
retriever = weaviate_pack.retriever
nodes = retriever.retrieve("query_str")

# use the query engine
query_engine = weaviate_pack.query_engine
response = query_engine.query("query_str")