Weaviate Vector Store

If you're opening this Notebook on colab, you will probably need to install LlamaIndex 🦙.

python

%pip install llama-index-vector-stores-weaviate

python

!pip install llama-index

Creating a Weaviate Client

python

import os
import openai

os.environ["OPENAI_API_KEY"] = ""
openai.api_key = os.environ["OPENAI_API_KEY"]

python

import logging
import sys

logging.basicConfig(stream=sys.stdout, level=logging.INFO)
logging.getLogger().addHandler(logging.StreamHandler(stream=sys.stdout))

python

import weaviate

python

# cloud
cluster_url = ""
api_key = ""

client = weaviate.connect_to_wcs(
    cluster_url=cluster_url,
    auth_credentials=weaviate.auth.AuthApiKey(api_key),
)

# local
# client = connect_to_local()

Load documents, build the VectorStoreIndex

python

from llama_index.core import VectorStoreIndex, SimpleDirectoryReader
from llama_index.vector_stores.weaviate import WeaviateVectorStore
from IPython.display import Markdown, display

Download Data

python

!mkdir -p 'data/paul_graham/'
!wget 'https://raw.githubusercontent.com/run-llama/llama_index/main/docs/examples/data/paul_graham/paul_graham_essay.txt' -O 'data/paul_graham/paul_graham_essay.txt'

python

# load documents
documents = SimpleDirectoryReader("./data/paul_graham").load_data()

python

from llama_index.core import StorageContext

# If you want to load the index later, be sure to give it a name!
vector_store = WeaviateVectorStore(
    weaviate_client=client, index_name="LlamaIndex"
)
storage_context = StorageContext.from_defaults(vector_store=vector_store)
index = VectorStoreIndex.from_documents(
    documents, storage_context=storage_context
)

# NOTE: you may also choose to define a index_name manually.
# index_name = "test_prefix"
# vector_store = WeaviateVectorStore(weaviate_client=client, index_name=index_name)

Using a custom batch configuration

Llamaindex defaults to Weaviate's dynamic batching, optimized for most common scenarios. However, in low-latency setups, this can overload the server or max out any GRPC Message limits in place. For more control and a better ingestion process, consider adjusting batch size by using the fixed size batch.

Here is how you can fine tune WeaviateVectorStore and define a custom batch:

python

from weaviate.classes.config import ConsistencyLevel

custom_batch = client.batch.fixed_size(
    batch_size=123,
    concurrent_requests=3,
    consistency_level=ConsistencyLevel.ALL,
)
vector_store_fixed = WeaviateVectorStore(
    weaviate_client=client,
    index_name="LlamaIndex",
    # we pass our custom batch as a client_kwargs
    client_kwargs={"custom_batch": custom_batch},
)

Query Index

python

# set Logging to DEBUG for more detailed outputs
query_engine = index.as_query_engine()
response = query_engine.query("What did the author do growing up?")

python

display(Markdown(f"<b>{response}</b>"))

Loading the index

Here, we use the same index name as when we created the initial index. This stops it from being auto-generated and allows us to easily connect back to it.

python

cluster_url = ""
api_key = ""

client = weaviate.connect_to_wcs(
    cluster_url=cluster_url,
    auth_credentials=weaviate.auth.AuthApiKey(api_key),
)

# local
# client = weaviate.connect_to_local()

python

vector_store = WeaviateVectorStore(
    weaviate_client=client, index_name="LlamaIndex"
)

loaded_index = VectorStoreIndex.from_vector_store(vector_store)

python

# set Logging to DEBUG for more detailed outputs
query_engine = loaded_index.as_query_engine()
response = query_engine.query("What happened at interleaf?")
display(Markdown(f"<b>{response}</b>"))

Metadata Filtering

Let's insert a dummy document, and try to filter so that only that document is returned.

python

from llama_index.core import Document

doc = Document.example()
print(doc.metadata)
print("-----")
print(doc.text[:100])

python

loaded_index.insert(doc)

python

from llama_index.core.vector_stores import ExactMatchFilter, MetadataFilters

filters = MetadataFilters(
    filters=[ExactMatchFilter(key="filename", value="README.md")]
)
query_engine = loaded_index.as_query_engine(filters=filters)
response = query_engine.query("What is the name of the file?")
display(Markdown(f"<b>{response}</b>"))

Deleting the index completely

You can delete the index created by the vector store using the delete_index function

python

vector_store.delete_index()

python

vector_store.delete_index()  # calling the function again does nothing

Connection Termination

You must ensure your client connections are closed:

python

client.close()