Releases: infiniflow/infinity
Releases · infiniflow/infinity
nightly
v0.4.0-dev2
Release v0.4.0-dev2 created from 14962eb at 2024-09-23 02:33:08+00:00
v0.4.0-dev1
Release v0.4.0-dev1 created from 81d7c7f at 2024-09-12 01:21:10+00:00
v0.3.0
v0.3.0-dev8
Release v0.3.0-dev8 created from 82b5e57 at 2024-09-02 10:01:19+00:00
infinity-v0.2.1
What's Changed
- Fix build script for nanobind by @yangzq50 in #1390
- Fix a docusaurus display issue by @writinwaters in #1393
- Fix: use static link for nanobind by @yangzq50 in #1392
- Remove redundant code by @JinHai-CN in #1396
- Rerank hnsw lvq by @small-turtle-1 in #1395
- Fix convert between embeddings by @Ognimalf in #1398
- Add export offset and limit by @JinHai-CN in #1399
- Opt bmp by @small-turtle-1 in #1400
- Add insert script for mldr benchmark by @yangzq50 in #1401
- Optimize memory by @Ma-cat in #1407
- Fix: data missing when flush block version file by @yangzq50 in #1406
- Install multiple python into builder image. by @yuzhichang in #1404
- Fix import error by @JinHai-CN in #1403
- Removed resource from rpm, deb and image by @yuzhichang in #1411
- Exact phrase optimization by @yuzhichang in #1412
- Export data into multiple files. by @JinHai-CN in #1409
- Accelerate release CI process by @JinHai-CN in #1413
- Create fulltext index optimization by @Ma-cat in #1410
- Update version information before release 0.2.1 by @JinHai-CN in #1414
- Update version by @JinHai-CN in #1420
- Fix path by @small-turtle-1 in #1421
- Fix: sort sparse expr in import and match statement by @small-turtle-1 in #1419
- python modules shall dynamicly link to libstdc++ by @yuzhichang in #1423
- Update version info to 0.2.1.dev3 by @JinHai-CN in #1424
- Add bpreorder by @small-turtle-1 in #1426
- Update MLDR benchmark script by @yangzq50 in #1425
- Fixed a docusaurus display issue by @writinwaters in #1430
- Update fusion weighted sum support and MLDR weighted sum benchmark by @yangzq50 in #1428
- build from source using clang-18 by @Ma-cat in #1429
- Fix parser by @small-turtle-1 in #1431
- Parallel execution of python test in ci by @Ma-cat in #1422
- Link to llvm libc++ instead of gcc libstdc++ by @JinHai-CN in #1435
- Clang18 by @yuzhichang in #1433
- Remove unused dockerfile and update benchmark record by @JinHai-CN in #1437
- Add script for generating colbert embeddings for MLDR benchmark by @yangzq50 in #1434
- Fix optimize index by @small-turtle-1 in #1436
- Name builder container with a random number to allow concurrent workflow by @yuzhichang in #1441
- Modify ci display by @Ma-cat in #1440
- support build infinity in ubuntu 24.04 by @Ma-cat in #1445
- Update version info by @JinHai-CN in #1439
- Add script for insert colbert embeddings for MLDR benchmark by @yangzq50 in #1442
- Optimize buffer mgr by @small-turtle-1 in #1444
- Limit cpu of the builder container and ninja by @yuzhichang in #1447
- Export CMAKE_BUILD_PARALLEL_LEVEL by @yuzhichang in #1450
- Fix txn manager bug by @small-turtle-1 in #1451
- Update MLDR benchmark script by @yangzq50 in #1449
- Add 'show logs' and 'show delta logs' by @JinHai-CN in #1453
- Add 'show catalogs' by @JinHai-CN in #1455
- fix create fulltext index by @Ma-cat in #1454
- Keep order of match expression by @yuzhichang in #1452
- Refactor meta data by @JinHai-CN in #1459
- Refactor entry_list by @JinHai-CN in #1443
- Fix formatting errors by @zjbpaul1317 in #1460
- Fix executor bug by @small-turtle-1 in #1458
- Optimize bp. by @small-turtle-1 in #1461
- Update mblas by @JinHai-CN in #1457
- Fix document errors by @zjbpaul1317 in #1464
- Update version info by @JinHai-CN in #1463
- Adapt slow_test for new runner by @yuzhichang in #1465
- Refactor meta_map by @JinHai-CN in #1462
- Update benchmark script by @JinHai-CN in #1467
- local infinity modify by @Ma-cat in #1468
- Fix slow_test and timeout_kill.sh for new runner by @yuzhichang in #1469
- Fix idxreplay by @small-turtle-1 in #1470
- Support local infinity to check the initialization path by @Ma-cat in #1471
- Fix fullckp by @small-turtle-1 in #1466
- Fix wrong meta directory by @JinHai-CN in #1472
- Fix SHOW and SET bugs by @JinHai-CN in #1474
- Update version and increase the fix length chunk size by @JinHai-CN in #1475
- Add: sparse benchmark. by @small-turtle-1 in #1476
- Fix update version info to 0.2.1 by @JinHai-CN in #1477
New Contributors
- @zjbpaul1317 made their first contribution in #1460
Full Changelog: v0.2.0...v0.2.1
v0.2.0
What's Changed
- Let thrift client reconnect on insert failure by @yuzhichang in #1156
- Declare fma requirement at CMakeLists.txt by @yuzhichang in #1162
- Update blockmax_maxscore_iterator.cppm by @eltociear in #1166
- Added python venv into CentOS 7 builder image by @yuzhichang in #1169
- Update infinity configs and variables. by @JinHai-CN in #1163
- Added quickstart guide. by @writinwaters in #1173
- Fix write conflict by @small-turtle-1 in #1170
- Fix path to build_from_source.md by @YReddice in #1186
- support phrase query with block max by @Ma-cat in #1189
- Support real time secondary index by @yangzq50 in #1171
- Rearranged docs by @writinwaters in #1190
- Fix unittest stuck by @small-turtle-1 in #1192
- Fix parallel readwrite by @small-turtle-1 in #1191
- optimize full-text index creation time by @Ma-cat in #1199
- correct python dependency installation command and add comments to gitignore by @sakura-mac in #1202
- Implement show configs, show session and global variables. by @JinHai-CN in #1203
- Initial support for tensor data type by @yangzq50 in #1205
- Refactor set command by @JinHai-CN in #1206
- Implement HTTP API: SHOW CONFIGS/VARIABLES/CONFIG/VARIABLE, SET CONFIG/VARIABLE by @JinHai-CN in #1207
- Fix python client connecting to server slowly. by @JinHai-CN in #1209
- Refactor tokenizer loading by @JinHai-CN in #1210
- Unified compaction. by @small-turtle-1 in #1208
- Fix MergeAggregate load meta by @yangzq50 in #1212
- Refactor benchmark script by @yuzhichang in #1213
- Updated benchmark.md by @yuzhichang in #1216
- Fix: compact parallel bug. by @small-turtle-1 in #1214
- Support Japanese morphological analyzer by @yingfeng in #1215
- Update build image for opencloudos and docs by @yingfeng in #1218
- Add MatchTensorScan operator by @yangzq50 in #1219
- LOG_ERROR message before raise recoverable error by @JinHai-CN in #1217
- Change log level by @JinHai-CN in #1222
- Add license of mecab by @yingfeng in #1223
- Refactor code issues by @JinHai-CN in #1224
- Refactor code by @JinHai-CN in #1225
- Refactor more codes by @JinHai-CN in #1226
- Refactor: update container and remove unused codes. by @JinHai-CN in #1227
- Introduced blockmax wand by @yuzhichang in #1229
- Merge duplicated TermQueryNode children by @yuzhichang in #1231
- Fix: compact bug. by @small-turtle-1 in #1230
- Fulltext preformance improvement by @yuzhichang in #1234
- Add support for TensorArray data type by @yangzq50 in #1233
- Fix: restart conflict by @yangzq50 in #1237
- Support default value and json format import for Tensor and TensorArray data type by @yangzq50 in #1241
- Simple sparse idx by @small-turtle-1 in #1238
- Optimize index creation time by @Ma-cat in #1240
- Refactor benchmark scripts by @yuzhichang in #1243
- Provide Tensor MaxSim reranker for Fusion operator by @yangzq50 in #1244
- Added back multiprocessing to benchmark scripts by @yuzhichang in #1246
- [Feature]: Exports table to JSONL or CSV file. by @JinHai-CN in #1245
- Removed MemoryPool and RecyclePool by @yuzhichang in #1248
- Check python SDK version when client connected with server. by @JinHai-CN in #1247
- Simple Sparse vector support. by @small-turtle-1 in #1249
- Minor improvement of fulltext query by @yuzhichang in #1250
- Add more support to sparse vector. by @small-turtle-1 in #1251
- Add more support to sparse vector. by @small-turtle-1 in #1254
- Refactor: Error logging by @JinHai-CN in #1256
- Fix: sparse info deserialize bug. by @small-turtle-1 in #1259
- Enable Phrase only for STANDARD by @yuzhichang in #1261
- Update python and http api for match tensor and fusion by @yangzq50 in #1255
- Fix sparse parser. by @small-turtle-1 in #1262
- Add distance and similarity metric as output in KNN search by @JinHai-CN in #1260
- Update client version by @JinHai-CN in #1264
- Fulltext benchmark improvement by @yuzhichang in #1263
- Support MatchTensor Scan and Reranker on TensorArray column by @yangzq50 in #1267
- Update Documents by @JinHai-CN in #1268
- fulltext Share PostingFormat among PostingWriters by @yuzhichang in #1270
- Update links by @writinwaters in #1271
- Add new command: show index segment info by @JinHai-CN in #1273
- Add new command: show index chunk info by @JinHai-CN in #1275
- Fix redundant code by @JinHai-CN in #1277
- Fix missing first_delete_ts_ by @yangzq50 in #1278
- Refactor code by @JinHai-CN in #1280
- Fixed query_batch type mismatch by @yuzhichang in #1282
- Sparse knn1 by @small-turtle-1 in #1283
- Add python example for ColBERT reranker by @yangzq50 in #1284
- Added rrf window_size by @yuzhichang in #1288
- Fix python scripts by @yangzq50 in #1287
- Multiway fusion by @yuzhichang in #1289
- Add three global variables by @JinHai-CN in #1290
- Refactor physical match by @JinHai-CN in #1285
- Fix read file function by @JinHai-CN in #1291
- Fix export embedding to wrong json type by @JinHai-CN in #1293
- [Bug] Fix import jsonl and json file, row count zero by @JinHai-CN in #1294
- Purge bad entries after latest checkpoint by @yuzhichang in #1296
- Stream import by @JinHai-CN in #1298
- Add show buffer command by @JinHai-CN in #1299
- Added weighted_sum fusion by @yuzhichang in #1301
- Changed weighted_sum normalization method by @yuzhichang in #1302
- Sparse knn2 by @small-turtle-1 in #1292
- Escape reserved character in query string by @yuzhichang in #1303
- Fix: mem leak. by @small-turtle-1 in #1304
- Fix garbage outputs of Chinese analyzer by @yingfeng in #1306
- API: export table to file by @JinHai-CN in #1300
- Add EMVB search: Part 1 by @yangzq50 in #1305
- Secondary index support string by @wuxiaobai24 in #1307
- Introduced CutGrain for Chinese analyzer by @yuzhichang in #1309
- Improve Chinese analyzer by @yingfeng in https://github.com/infiniflow/infinity/p...
infinity-v0.1.1
What's Changed
- Update benchmark.md by @writinwaters in #1158
- Fix benchmark doc error by @JinHai-CN in #1159
- Add specific pysdk version to README by @JinHai-CN in #1160
- Editorial updates to RESTful APIs by @writinwaters in #1165
- Added quickstart by @writinwaters in #1168
- Support real time secondary index by @yangzq50 in #1187
- Fix write conflict (#1170) by @small-turtle-1 in #1188
- Fix bug in 0.1.x by @small-turtle-1 in #1195
- Prepare for release 0.1.1 by @JinHai-CN in #1196
infinity-v0.1.0
Infinity v0.1.0 is the first release that encapsulates our initial definition for an AI-native database: capable of performing searches or combined searches of any of the following categories:
- Structured data query
- Embedding/vector search
- Full-text search
Installation
Clients
Data types
See here for more information.
Benchmark
See the benchmark report.
Key features
See here for more information.