Release 2.4.11

What's Changed

• Delete some third-party dependencies in #222 #226 #236
• Support Android @Presburger in #229
• Support IVF_FLAT index by @junjiejiangjjj in #234
  After creating the IVF_FLAT index, the index will be built when the data exceeds 100,000
  example:

from pymilvus import MilvusClient, DataType

# 1. Set up a Milvus client
client = MilvusClient(
    uri="./milvus_demo.db"
)

# 2. Create schema
# 2.1. Create schema
schema = MilvusClient.create_schema(
    auto_id=False,
    enable_dynamic_field=True,
)

# 2.2. Add fields to schema
schema.add_field(field_name="id", datatype=DataType.INT64, is_primary=True)
schema.add_field(field_name="vector", datatype=DataType.FLOAT_VECTOR, dim=5)

# 3. Create collection
client.create_collection(
    collection_name="customized_setup", 
    schema=schema, 
)
# 4.1. Set up the index parameters
index_params = MilvusClient.prepare_index_params()

# 4.2. Add an index on the vector field.
index_params.add_index(
    field_name="vector",
    metric_type="COSINE",
    index_type="IVF_FLAT",
    index_name="vector_index",
    params={ "nlist": 128 }
)

# 4.3. Create an index file
client.create_index(
    collection_name="customized_setup",
    index_params=index_params
)

• Fix some bugs

New Contributors

• @982945902 made their first contribution in #224

Full Changelog: v2.4.10...v2.4.11