llama-index vector_stores lindorm integration
Project description
LlamaIndex Vector_Stores Integration: Lindorm
- LindormVectorStore support pure vector search, search with metadata filtering, hybrid search, async, etc.
- Please refer to the notebook for usage of Lindorm as vector store in LlamaIndex.
Example Usage
pip install llama-index
pip install opensearch-py
pip install llama-index-vector-stores-lindorm
from llama_index.vector_stores.lindorm import (
LindormVectorStore,
LindormVectorClient,
)
# how to obtain an lindorm search instance:
# https://alibabacloud.com/help/en/lindorm/latest/create-an-instance
# how to access your lindorm search instance:
# https://www.alibabacloud.com/help/en/lindorm/latest/view-endpoints
# run curl commands to connect to and use LindormSearch:
# https://www.alibabacloud.com/help/en/lindorm/latest/connect-and-use-the-search-engine-with-the-curl-command
# lindorm instance info
host = "ld-bp******jm*******-proxy-search-pub.lindorm.aliyuncs.com"
port = 30070
username = "your_username"
password = "your_password"
# index to demonstrate the VectorStore impl
index_name = "lindorm_test_index"
# extension param of lindorm search, number of cluster units to query; between 1 and method.parameters.nlist.
nprobe = "a number(string type)"
# extension param of lindorm search, usually used to improve recall accuracy, but it increases performance overhead;
# between 1 and 200; default: 10.
reorder_factor = "a number(string type)"
# LindormVectorClient encapsulates logic for a single index with vector search enabled
client = LindormVectorClient(
host=host,
port=port,
username=username,
password=password,
index=index_name,
dimension=1536, # match with your embedding model
nprobe=nprobe,
reorder_factor=reorder_factor,
# filter_type="pre_filter/post_filter(default)"
)
# initialize vector store
vector_store = LindormVectorStore(client)
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Close
Hashes for llama_index_vector_stores_lindorm-0.2.0.tar.gz
Algorithm | Hash digest | |
---|---|---|
SHA256 | a0d224cc0ba043f329f60fc1bb736eeca7e5b992ef789aa3bc80a8b6ca2f9bb6 |
|
MD5 | 0619cf0cc6ee2afc7a512a6454d6299b |
|
BLAKE2b-256 | 914bbb540f7cefd0576e917061fb16a4115f39949ce1a9c96aef3d8ac691d75d |
Close
Hashes for llama_index_vector_stores_lindorm-0.2.0-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | d118cebbdde3b688edbd29060ea59967a6599cd84e5d7f902d5b9a05eaf27c61 |
|
MD5 | fa98735ea95bd2ef94788ef4c7563272 |
|
BLAKE2b-256 | 24cad1400a7b5116a8d859e2e9f0e4c4fb438e0c3e1ec03784323b35788d0834 |