Commit Graph

17 Commits (350dde666dc1d99f6d5a257b595b2be5d74b3fda)

Author SHA1 Message Date
zhagnlu 4d2f96c760
enhance: support bitmap mmap (#35399)
#32900

Signed-off-by: luzhang <luzhang@zilliz.com>
Co-authored-by: luzhang <luzhang@zilliz.com>
2024-08-27 16:34:59 +08:00
zhenshan.cao aa247f192d
enhance: remove unused code for StorageV2 (#35132)
issue: https://github.com/milvus-io/milvus/issues/34168

Signed-off-by: zhenshan.cao <zhenshan.cao@zilliz.com>
2024-08-01 12:08:13 +08:00
zhagnlu d43ec4db0b
enhance: support array bitmap index (#33527)
#32900

---------

Signed-off-by: luzhang <luzhang@zilliz.com>
Co-authored-by: luzhang <luzhang@zilliz.com>
2024-06-16 21:51:58 +08:00
Jiquan Long 0c5d8660aa
feat: support inverted index for array (#33452)
issue: https://github.com/milvus-io/milvus/issues/27704

---------

Signed-off-by: longjiquan <jiquan.long@zilliz.com>
2024-05-31 09:47:47 +08:00
zhagnlu 589d4dfd82
enhance: optimize bitmap index (#33358)
#32900

Signed-off-by: luzhang <luzhang@zilliz.com>
Co-authored-by: luzhang <luzhang@zilliz.com>
2024-05-30 13:09:43 +08:00
zhagnlu d669fbcf46
enhance: support bitmap index for scalar type (#32902)
#32900

Signed-off-by: luzhang <luzhang@zilliz.com>
Co-authored-by: luzhang <luzhang@zilliz.com>
2024-05-19 21:49:38 +08:00
Buqian Zheng 96cfae55a5
feat: [Sparse Float Vector] segcore to support sparse vector search and get raw vector by id (#30629)
This PR adds the ability to search/get sparse float vectors in segcore,
and added unit tests by modifying lots of existing tests into
parameterized ones.

https://github.com/milvus-io/milvus/issues/29419

Signed-off-by: Buqian Zheng <zhengbuqian@gmail.com>
2024-03-12 09:16:30 -07:00
zhagnlu 976b6fc0e4
enhance: change opendal as compile configurable (#30384)
#30373

Signed-off-by: luzhang <luzhang@zilliz.com>
Co-authored-by: luzhang <luzhang@zilliz.com>
2024-02-20 19:16:52 +08:00
Jiquan Long 3f46c6d459
feat: support inverted index (#28783)
issue: https://github.com/milvus-io/milvus/issues/27704

Add inverted index for some data types in Milvus. This index type can
save a lot of memory compared to loading all data into RAM and speed up
the term query and range query.

Supported: `INT8`, `INT16`, `INT32`, `INT64`, `FLOAT`, `DOUBLE`, `BOOL`
and `VARCHAR`.

Not supported: `ARRAY` and `JSON`.

Note:
- The inverted index for `VARCHAR` is not designed to serve full-text
search now. We will treat every row as a whole keyword instead of
tokenizing it into multiple terms.
- The inverted index don't support retrieval well, so if you create
inverted index for field, those operations which depend on the raw data
will fallback to use chunk storage, which will bring some performance
loss. For example, comparisons between two columns and retrieval of
output fields.

The inverted index is very easy to be used.

Taking below collection as an example:

```python
fields = [
		FieldSchema(name="pk", dtype=DataType.VARCHAR, is_primary=True, auto_id=False, max_length=100),
		FieldSchema(name="int8", dtype=DataType.INT8),
		FieldSchema(name="int16", dtype=DataType.INT16),
		FieldSchema(name="int32", dtype=DataType.INT32),
		FieldSchema(name="int64", dtype=DataType.INT64),
		FieldSchema(name="float", dtype=DataType.FLOAT),
		FieldSchema(name="double", dtype=DataType.DOUBLE),
		FieldSchema(name="bool", dtype=DataType.BOOL),
		FieldSchema(name="varchar", dtype=DataType.VARCHAR, max_length=1000),
		FieldSchema(name="random", dtype=DataType.DOUBLE),
		FieldSchema(name="embeddings", dtype=DataType.FLOAT_VECTOR, dim=dim),
]
schema = CollectionSchema(fields)
collection = Collection("demo", schema)
```

Then we can simply create inverted index for field via:

```python
index_type = "INVERTED"
collection.create_index("int8", {"index_type": index_type})
collection.create_index("int16", {"index_type": index_type})
collection.create_index("int32", {"index_type": index_type})
collection.create_index("int64", {"index_type": index_type})
collection.create_index("float", {"index_type": index_type})
collection.create_index("double", {"index_type": index_type})
collection.create_index("bool", {"index_type": index_type})
collection.create_index("varchar", {"index_type": index_type})
```

Then, term query and range query on the field can be speed up
automatically by the inverted index:

```python
result = collection.query(expr='int64 in [1, 2, 3]', output_fields=["pk"])
result = collection.query(expr='int64 < 5', output_fields=["pk"])
result = collection.query(expr='int64 > 2997', output_fields=["pk"])
result = collection.query(expr='1 < int64 < 5', output_fields=["pk"])
```

---------

Signed-off-by: longjiquan <jiquan.long@zilliz.com>
2023-12-31 19:50:47 +08:00
cai.zhang fb089cda8b
enhance: Load raw data while scalar index doesn't have raw data (#28888)
issue: #28886

Signed-off-by: Cai Zhang <cai.zhang@zilliz.com>
2023-12-06 20:36:36 +08:00
Bingyi Sun 36f69ea031
feat: integrate storagev2 in building index of segcore (#28768)
issue: https://github.com/milvus-io/milvus/issues/28655

---------

Signed-off-by: sunby <sunbingyi1992@gmail.com>
2023-12-05 16:48:54 +08:00
Jiquan Long a36fefb009
Fix cpplint (#22657)
Signed-off-by: longjiquan <jiquan.long@zilliz.com>
2023-03-10 09:47:54 +08:00
presburger 9950cacd10
support knowhere 2.0 (#21857)
Signed-off-by: Yusheng.Ma <Yusheng.Ma@zilliz.com>
2023-02-10 14:24:32 +08:00
xige-16 428840178c
Support diskann index for vector field (#19093)
Signed-off-by: xige-16 <xi.ge@zilliz.com>

Signed-off-by: xige-16 <xi.ge@zilliz.com>
2022-09-21 20:16:51 +08:00
xige-16 56778787be
Reverse data from scalar index (#17145)
Signed-off-by: xige-16 <xi.ge@zilliz.com>
2022-05-26 14:58:01 +08:00
xige-16 515d0369de
Support string type in segcore (#16546)
Signed-off-by: xige-16 <xi.ge@zilliz.com>
Co-authored-by: dragondriver <jiquan.long@zilliz.com>

Co-authored-by: dragondriver <jiquan.long@zilliz.com>
2022-04-29 13:35:49 +08:00
Jiquan Long fd589baca7
Integrates marisa trie index (#16192)
Signed-off-by: dragondriver <jiquan.long@zilliz.com>
2022-04-01 15:31:29 +08:00