PyMilvus 2.3.0 Release Notes

What's Changed

• Add the load_state api by @SimFG in #1258
• add resource group api by @weiliu1031 in #1274
• Support rename collection API by @jaime0815 in #1299
• Support upsert by @smellthemoon in #1303
• Add ignore_growing option for query and search by @aoiasd in #1286
• Add GPU index types by @cydrain in #1314
• Add RAFT GPU index types by @cydrain in #1323
• Add FlushAll api by @bigsheeper in #1334
• Add new metric type COSINE by @cydrain in #1376
• Add an utility to get server type by @longjiquan in #1381
• List indexes by field name by @longjiquan in #1386
• Remove default of shards_num for new collection by @XuanYang-cn in #1387
• Milvus Client by @filip-halt in #1374
• Support insert dict as JSON by @yah01 in #1424
• Support Partition key by @xige-16 in #1427
• support iterator for query and search by @MrPresent-Han in #1431
• Add support of insert by rows by @czs007 in #1434
• Support api key by @bigsheeper in #1460
• Call Connect after db name changed by @longjiquan in #1484
• Pass the is_dynamic field to the milvus server by @xiaocai2333 in #1495
• Support return pending index rows when describe index by @xiaocai2333 in #1504
• Do not set metric_type default value in search params (#1469) by @bigsheeper in #1511
• Support async refresh by @yah01 in #1537
• Change pylint to ruff and black by @XuanYang-cn in #1565
• add autoID to varchar dataType by @smellthemoon in #1583
• Support offset in search param by @XuanYang-cn in #1613
• Remove time travel params by @jiaoew1991 in #1635
• Disable logger from polluting other modules by @ponponon in #1646

New Contributors

• @liliu-z made their first contribution in #1223
• @smellthemoon made their first contribution in #1303
• @leykun10 made their first contribution in #1338
• @MrPresent-Han made their first contribution in #1431
• @pingliu made their first contribution in #1520
• @im-ajaymeena made their first contribution in #1517
• @shunjiezhao made their first contribution in #1551
• @slobentanzer made their first contribution in #1649
• @jyc4617 made their first contribution in #1638
• @ponponon made their first contribution in #1646

Full Changelog: v2.2.15...v2.3.0

PyMilvus 2.2.15 Release Notes

1. Fixed some bugs.
2. Enable to set offet&limit in search params.

PyMilvus 2.2.14 Release Notes

1. Support rename db for collection
2. Support describe index with timestamp
3. Update bulkinsert examples
4. Fix some bugs

PyMilvus 2.2.13 Release Notes

Fix apache software license classifier (#1563)

Signed-off-by: zhenshan.cao <[email protected]>

PyMilvus 2.2.12 Release Notes

Fixing the encoding issue

By setting the "ensure_ascii=False" parameter in the "json.dumps" function, non-ASCII characters in the data are not escaped during the "insert" operation.
After inserting non-ASCII characters without escaping, users can also use them as keys in expression filtering from the JSON field.
The following is a sample code snippet.

import numpy as np
from pymilvus import MilvusClient, DataType

dimension = 128
collection_name = "books"
client = MilvusClient("http://localhost:19530")

schema = client.create_schema(auto_id=True)
schema.add_field("id", DataType.INT64, is_primary=True)
schema.add_field("embeddings", DataType.FLOAT_VECTOR, dim=dimension)
schema.add_field("info", DataType.JSON)

index_param = client.prepare_index_params("embeddings", metric_type="L2")
client.create_collection_with_schema(collection_name, schema, index_param)

rng = np.random.default_rng(seed=19530)
rows = [
    {"embeddings": rng.random((1, dimension))[0],
     "info": {"title": "Lord of the Flies", "author": "William Golding"}},

    {"embeddings": rng.random((1, dimension))[0],
     "info": {"作者": "J.D.塞林格", "title": "麦田里的守望者", }},

    {"embeddings": rng.random((1, dimension))[0],
     "info": {"Título": "Cien años de soledad", "autor": "Gabriel García Márquez"}},
]

client.insert(collection_name, rows)
result = client.query(collection_name, filter="info['作者'] == 'J.D.塞林格' or info['Título'] == 'Cien años de soledad'",
                      output_fields=["info"],
                      consistency_level="Strong")

for hit in result:
    print(f"hit: {hit}")

The output will be:

hit: {'info': {'作者': 'J.D.塞林格', 'title': '麦田里的守望者'}, 'id': 442210659570062545}
hit: {'info': {'Título': 'Cien años de soledad', 'autor': 'Gabriel García Márquez'}, 'id': 442210659570062546}

PyMilvus 2.2.11 Release Notes

Add milvus_client

PyMilvus 2.2.9 Release Notes

• JSON support
• Dynamic schema
• Partition key
• Connection management

PyMilvus 2.2.8 Release Notes

New Features

1. [FEATURE]: Has an utility function to parse if the connection is to Zilliz cloud or opensource Milvus

>>> utility.get_server_type(using="default")
"milvus"

2. List indexes by field name #1386

>>> utility.list_indexes(collection_name, field_name=vec_field)
[vec_field_idx]

3. [FEATURE]: Remove the logic to fill in the default number of shard when create collection
   PyMilvus doesn't set default shard numbers when creating a new collection.
4. [FEATURE]: PyMilvus supports using env configs

• Support uri with username and password:

from pymilvus import connections
uri = "https://username:[email protected]:19530"
connections.connect(uri=uri)

• Support using MIVLUS_URI env to init default connection alias.

// set MILVUS_URI in env
$ export MILVUS_URI=https://username:[email protected]:19530

>>> from pymilvus import connections
>>> connections.connect()
>>> connections.get_connection_addr("default")
{"address": "exampledomain.com:19530", "user": username}

• Support reading envs from .env file

# .env.example in https://github.com/milvus-io/pymilvus/blob/master/.env.example
# Please copy this file and rename as .env, pymilvus will read .env file if provided

MILVUS_URI=
# MILVUS_URI=https://username:[email protected]:19530

# Milvus connections configs
MILVUS_CONN_ALIAS=default
MILVUS_CONN_TIMEOUT=10

Bug fixes

• [Bug]: Partition in orm is buggy

PyMilvus 2.2.7 Release Notes

Fix some bugs

PyMilvus 2.2.6 Release Notes

Fix some bugs