{"author":"Milvus Team","author_email":"milvus-team@zilliz.com","bugtrack_url":null,"classifiers":[],"description":"

\n\n

\n
A lightweight version of Milvus
\n

\n\n

\n \n $\"license\"/$ \n \n \n $\"github$ \n \n \n $\"github$ \n \n

\n\n\n# Introduction\nMilvus Lite is the lightweight version of [Milvus](https://github.com/milvus-io/milvus), a high-performance vector database that powers AI applications with vector similarity search. This repo contains the core components of Milvus Lite. \n\nWith Milvus Lite, you can start building an AI application with vector similarity search within minutes! Milvus Lite is good for running in the following environment:\n- Jupyter Notebook / Google Colab\n- Laptops\n- Edge Devices\n\nMilvus Lite can be imported into your Python application, providing the core vector search functionality of Milvus. Milvus Lite is already included in the [Python SDK of Milvus](https://github.com/milvus-io/pymilvus). To use it, you just need `pip install pymilvus`. \n\nMilvus Lite uses the same API as Milvus Standalone and Distributed, providing a consistent experience across environments. Develop your GenAI applications once and run them anywhere: on a laptop or Jupyter Notebook with Milvus Lite, in a Docker container with Milvus Standalone, or on a K8s cluster with Milvus Distributed for large-scale production.\n\n

\n\nMilvus Lite is only suitable for small scale prototyping (usually less than a million vectors) or edge devices. For large scale production, we recommend using [Milvus Standalone](https://milvus.io/docs/install-overview.md#Milvus-Standalone) or [Milvus Distributed](https://milvus.io/docs/install-overview.md#Milvus-Distributed). You can also consider the fully-managed Milvus on [Zilliz Cloud](https://zilliz.com/cloud).\n\n# Requirements\nMilvus Lite currently supports the following environments:\n- Ubuntu >= 20.04 (x86_64 and arm64)\n- MacOS >= 11.0 (Apple Silicon M1/M2 and x86_64)\n\n***Note:*** Windows is not yet supported.\n\n# Installation\n```shell\npip install -U pymilvus\n```\nWe recommend using `pymilvus`. Since `milvus-lite` is included in `pymilvus` version 2.4.2 or above, you can `pip install` with `-U` to force update to the latest version and `milvus-lite` is automatically installed.\n\n\nIf you want to explicitly install `milvus-lite` package, or you have installed an older version of `milvus-lite` and would like to update it, you can do `pip install -U milvus-lite`.\n\n# Usage\nIn `pymilvus`, specify a local file name as uri parameter of MilvusClient will use Milvus Lite.\n```python\nfrom pymilvus import MilvusClient\nclient = MilvusClient(\"./milvus_demo.db\")\n```\n\n> **_NOTE:_** Note that the same API also applies to Milvus Standalone, Milvus Distributed and Zilliz Cloud, the only difference is to replace local file name to remote server endpoint and credentials, e.g. \n`client = MilvusClient(uri=\"http://localhost:19530\", token=\"username:password\")` for self-hosted Milvus server.\n\n# Examples\nFollowing is a simple demo showing how to use Milvus Lite for text search. There are more comprehensive [examples](https://github.com/milvus-io/bootcamp/tree/master/bootcamp/tutorials) for using Milvus Lite to build applications\nsuch as [RAG](https://github.com/milvus-io/bootcamp/blob/master/bootcamp/tutorials/quickstart/build_RAG_with_milvus.ipynb), [image search](https://github.com/milvus-io/bootcamp/blob/master/bootcamp/tutorials/quickstart/image_search_with_milvus.ipynb), and using Milvus Lite in popular RAG framework such as [LangChain](https://github.com/milvus-io/bootcamp/blob/master/bootcamp/tutorials/integration/rag_with_milvus_and_langchain.ipynb) and [LlamaIndex](https://github.com/milvus-io/bootcamp/blob/master/bootcamp/tutorials/integration/rag_with_milvus_and_llamaindex.ipynb)!\n\n```python\nfrom pymilvus import MilvusClient\nimport numpy as np\n\nclient = MilvusClient(\"./milvus_demo.db\")\nclient.create_collection(\n collection_name=\"demo_collection\",\n dimension=384 # The vectors we will use in this demo has 384 dimensions\n)\n\n# Text strings to search from.\ndocs = [\n \"Artificial intelligence was founded as an academic discipline in 1956.\",\n \"Alan Turing was the first person to conduct substantial research in AI.\",\n \"Born in Maida Vale, London, Turing was raised in southern England.\",\n]\n# For illustration, here we use fake vectors with random numbers (384 dimension).\n\nvectors = [[ np.random.uniform(-1, 1) for _ in range(384) ] for _ in range(len(docs)) ]\ndata = [ {\"id\": i, \"vector\": vectors[i], \"text\": docs[i], \"subject\": \"history\"} for i in range(len(vectors)) ]\nres = client.insert(\n collection_name=\"demo_collection\",\n data=data\n)\n\n# This will exclude any text in \"history\" subject despite close to the query vector.\nres = client.search(\n collection_name=\"demo_collection\",\n data=[vectors[0]],\n filter=\"subject == 'history'\",\n limit=2,\n output_fields=[\"text\", \"subject\"],\n)\nprint(res)\n\n# a query that retrieves all entities matching filter expressions.\nres = client.query(\n collection_name=\"demo_collection\",\n filter=\"subject == 'history'\",\n output_fields=[\"text\", \"subject\"],\n)\nprint(res)\n\n# delete\nres = client.delete(\n collection_name=\"demo_collection\",\n filter=\"subject == 'history'\",\n)\nprint(res)\n```\n\n# Supported Features\n\n## Functionality\nMilvus Lite shares the same API as Milvus Standalone, Milvus Distributed and Zilliz Cloud, offering core features including:\n- Insert/upsert operations\n- Vector data persistence and collection management\n- Dense, sparse, and hybrid vector search\n- Metadata filtering\n- Multi-vector support \n\n## Index Type\nMilvus Lite has limited index type support compared to other Milvus deployments:\n\n- Prior to version 2.4.11:\n - Only supports the [FLAT](https://milvus.io/docs/index.md?tab=floating#FLAT) index type\n - Uses FLAT index regardless of the specified index type in collection creation\n\n- Version 2.4.11 and later:\n - Supports both [FLAT](https://milvus.io/docs/index.md?tab=floating#FLAT) and [IVF_FLAT](https://milvus.io/docs/index.md?tab=floating#IVFFLAT) index types\n - For IVF_FLAT indexes:\n - Data size < 100,000: Automatically uses FLAT index internally for better performance\n - Data size ≥ 100,000: Constructs and uses IVF_FLAT index as specified\n\n# Known Limitations\n\nMilvus Lite does not support partitions, users/roles/RBAC, alias. To use those features, please choose other Milvus deployment types such as [Standalone](https://milvus.io/docs/install-overview.md#Milvus-Standalone), [Distributed](https://milvus.io/docs/install-overview.md#Milvus-Distributed) or [Zilliz Cloud](https://zilliz.com/cloud) (fully-managed Milvus).\n\n# Migrating data from Milvus Lite\n\nAll data stored in Milvus Lite can be easily exported and loaded into other types of Milvus deployment, such as Milvus Standalone on Docker, Milvus Distributed on K8s, or fully-managed Milvus on [Zilliz Cloud](https://zilliz.com/cloud).\n\nMilvus Lite provides a command line tool that can dump data into a json file, which can be imported into self-hosted [Milvus](https://github.com/milvus-io/milvus) or fully-managed Milvus on [Zilliz Cloud](https://zilliz.com/cloud). \n\n```shell\npip install -U \"pymilvus[bulk_writer]\"\n# The `milvus-lite` command line tool is already included in `milvus-lite` package which is part of \"pymilvus\", but it also needs some dependencies from `pymilvus[bulk_writer]` for dumping data.\n\nmilvus-lite dump -h\n\nusage: milvus-lite dump [-h] [-d DB_FILE] [-c COLLECTION] [-p PATH]\n\noptional arguments:\n -h, --help show this help message and exit\n -d DB_FILE, --db-file DB_FILE\n milvus lite db file\n -c COLLECTION, --collection COLLECTION\n collection that need to be dumped\n -p PATH, --path PATH dump file storage dir\n```\nThe following example dumps all data from `demo_collection` collection that's stored in `./milvus_demo.db` (a user specified local file that persists data for Milvus Lite)\n\nTo export data:\n\n```shell\nmilvus-lite dump -d ./milvus_demo.db -c demo_collection -p ./data_dir\n# ./milvus_demo.db: milvus lite db file\n# demo_collection: collection that need to be dumped\n#./data_dir : dump file storage dir\n```\n\nYou can use the dump file as input to upload data to Zilliz Cloud via [Data Import](https://docs.zilliz.com/docs/data-import), or a self-hosted Milvus server via [Bulk Insert](https://milvus.io/docs/import-data.md).\n# Contributing\nIf you want to contribute to Milvus Lite, please read the [Contributing Guide](https://github.com/milvus-io/milvus-lite/blob/main/CONTRIBUTING.md) first.\n\n# Report a bug\nFor any bug or feature request, please report it by submitting an issue in [milvus-lite](https://github.com/milvus-io/milvus-lite/issues/new/choose) repo.\n\n# License\nMilvus Lite is under the Apache 2.0 license. See the [LICENSE](https://github.com/milvus-io/milvus-lite/blob/main/LICENSE) file for details.\n","description_content_type":"text/markdown","docs_url":null,"download_url":null,"downloads":{"last_day":-1,"last_month":-1,"last_week":-1},"dynamic":null,"home_page":"https://github.com/milvus-io/milvus-lite.git","keywords":null,"license":null,"license_expression":null,"license_files":null,"maintainer":null,"maintainer_email":null,"name":"milvus-lite","package_url":"https://pypi.org/project/milvus-lite/","platform":null,"project_url":"https://pypi.org/project/milvus-lite/","project_urls":{"Homepage":"https://github.com/milvus-io/milvus-lite.git"},"provides_extra":null,"release_url":"https://pypi.org/project/milvus-lite/2.4.12/","requires_dist":["tqdm"],"requires_python":">=3.7","summary":"A lightweight version of Milvus wrapped with Python.","version":"2.4.12","yanked":false,"yanked_reason":null}

\n A lightweight version of Milvus\n

\n
A lightweight version of Milvus
\n

Package search

milvus-lite

project_urls

Dependencies

All package metadata

Releases