Releases: eosphoros-ai/DB-GPT
v0.5.7
✨ Enhancements
- Support intent detection (#1588)
- Add BM25 Retriever. (#1578)
- Support openchat-3.6-8b-20240522 (#1564)
- Support llama-3-sqlcoder-8b (#1559)
- Support Phi-3 models (#1554)
- Add aload_document (#1548)
- Support Yi 1.5 16k models (#1543)
- Support DeepSeek v2 lite model (#1537)
- Add support for Vertica analytical database (#1538)
🐞 Bug fixes
- Fix cannot connect other vector store (#1579)
- Update data_loader.py - timestamp+id (#1572)
- Fix storage connector error (#1547)
- Fix CrossEncoderRanker (#1531)
- Fix data dashboard - data_loader.py (#1526)
- Fix the problem of multiple line breaks in Q&A (#1573)
- Remove resources of planner agent (#1540)
- Wrong default value of parameter parallel_executions for async_retry (#1545)
🛠️ Other improvements
- update wechat (#1544)
Thank you to all our contributors for making this release possible!
@Aries-ckt, @FOkvj, @dingqiangliu, @fangyinc, @wzc570738205 and @yongzheJIN
v0.5.6
✨ Enhancements
- Support resource for dbgpts (#1527)
- ChatKnowledge Support Excel Document (#1524)
- Add GraphRAG framework and integrate TuGraph (#1506)
- Support deploy rerank model (#1522)
- Refactor resource of agents (#1518)
- Support Yi-1.5 models (#1516)
- Chroma store refactor (#1508)
- ES VectorStore (#1500)
- Support DeepSeek proxy LLM (#1491)
- Agent modular refactoring (#1487)
- Support ollama as an optional llm & embedding proxy (#1475)
- Support Qwen110B (#1474)
🐞 Bug fixes
- Fix Multi vector client error (#1532)
- Fix web build error (#1530)
- Fix build allinone images error (#1529)
- Fix the issue where the knowledge base cannot be edited when using ve… (#1519)
- Upgrade spacy model (#1511)
- Fix CrossEncoderRanker bug of EmbeddingRetriever (#1504)
- Fix rerank bug of EmbeddingRetriever (#1499)
- Fix app pydantic error (#1478)
- Fix Milvus will autoflush, manual flush is slowly (#1470)
🛠️ Other improvements
- Add ollama document (#1512)
Thank you to all our contributors for making this release possible!
@Aries-ckt, @FOkvj, @GITHUBear, @IamWWT, @JamesBonddu, @fangyinc, @fanzhidongyzby, @ivanzhu109, @jsRuner, @yyhhyyyyyy and @zhenchaozhu
v0.5.5
✨ Enhancements
- Added support for TuGraph graph database (#1451)
- Add MRR and HitRate retriever metrics. (#1456)
- Add CrossEncoder rerank (#1442)
- Support Qwen2MoE (#1439)
- Support Llama-3 (#1436)
- Support OceanBase vector database (#1435)
- Datasource support summary refresh (#1433)
- Upgrade pydantic to 2.x (#1428)
- Support CodeQwen1.5-7B-Chat (#1431)
- Support retry for 'Chat Data' (#1419)
- Dbgpts support agent (#1417)
- Support oceanbase monaco plugin (#1386)
🐞 Bug fixes
- Fix parameter type mismatch issue when using Milvus (#1449)
- Fix oceanbase vector_name_exists bug (#1465)
- Fix clickhouse occur error AttributeError: 'ClickhouseConnect… (#1432)
- Fix awel check for empty DataFrame data bug (#1430)
- Fix rdbms_db_summary (#1409)
Thank you to all our contributors for making this release possible!
@Aries-ckt, @GITHUBear, @HSunboy, @KingSkyLi, @dusens, @fangyinc, @jsRuner and @yyhhyyyyyy
v0.5.4
✨ Enhancements
- Modify default webserver port to 5670 (#1410)
- Add trace for agent (#1407)
- Support moonshot proxy LLM (#1404)
- Release agent SDK (#1396)
- Add metadata properties filters (#1395)
- Upgrade torch version to 2.2.1 (#1374)
- Support HTTP sender (#1383)
- Support Qwen1.5-32B (#1385)
- Support Starling-LM-7B-beta (#1350)
- Optimize dashboard UI and fix sql highlight (#1329)
🐞 Bug fixes
- Fix Code standards (#1393)
- Fix Resolve the issue of incorrect parameter echo for resources under app editing (#1391)
- Fix the problem of resource duplication and solve the problem of slow recall of table structure (#1381)
- Fix dashboard editor run error (#1380)
- Fix the mypy check error (#1373)
- Fix the default dependency installation error (#1363)
- Fix when delete knowledge list error embedding is required for Milvus (#1359)
- Fix the issue of invalid port settings (#1351)
🛠️ Other improvements
- Update readme content (#1392)
Thank you to all our contributors for making this release possible!
@Aries-ckt, @evanchenhi, @fangyinc, @fanzhidongyzby, @jsRuner, @lcxadml, @majianguang, @xiuzhu9527 and @yyhhyyyyyy
v0.5.3
✨ Enhancements
🐞 Bug fixes
🛠️ Other improvements
- Add AWEL chat data cookbook (#1345)
Thank you to all our contributors for making this release possible!
@Aries-ckt, @csunny and @fangyinc
v0.5.2
✨ Enhancements
- Support RAG SDK (#1322)
- Add Client and API v2 (#1316)
- Refactor datasource module (#1309)
- Support yi proxy LLM (#1303)
- Support rag retriever evaluation (#1291)
- Add support for multiple versions of the same model (#1246)
🐞 Bug fixes
- Fix sql format error (#1319)
- Fix awel check empty data bug (#1311)
- Fix the bug of doris missing comments (#1308)
- Fix ClickHouse default login bug (#1292)
- Fix agent loss message bug (#1283)
- Fixed the problem of loading only one plugin when there are multiple (#1282)
🛠️ Other improvements
Thank you to all our contributors for making this release possible!
@2976151305, @ASTLY123, @Aries-ckt, @fangyinc, @jsRuner, @lordk911 and @yhjun1026
v0.5.1
✨ Enhancements
- APIServer supports embeddings (#1256)
- AWEL flow supports dynamic parameters (#1251)
- Support Hive Conn (#1215)
- Dynamically loading dbgpts (#1211)
- Support Copy AWEL flow (#1200)
- Support default strategy batch sync. (#1196)
- Support gemma model (#1187)
- Support RAG chat flow (#1185)
🚀 Performance improvements
🐞 Bug fixes
- Fix multi-turn dialogue bug (#1259)
- Fix knowledge_space_dao.update_knowledge_space call in Knowledge… (#1250)
- Fix scheduler shutdown (#1237)
- Fix typo error (#1228)
- Fix HTTP Trigger parse parameter bug (#1225)
- Fix switch strategy failed in Chinese env. (#1219)
- Fix milvus_store.py delete collection(#1212)
- Fix KnowledgeQueryResponse (#1202)
- Fix sql format error (#1199)
🛠️ Other improvements
- Upgrate v0.5.1 doc (#1265)
- Overview update & fix reference addr (#1263)
- New AWEL tutorial (#1245)
- Set default knowledge icon (#1244)
- Add new feature docs reference (#1238)
- Add docs and usage for v0.5.0 (#1216)
- Update comments for gender and pay_status fields in test_case_m… (#1198)
Thank you to all our contributors for making this release possible!
@0sengseng0, @2976151305, @Aralhi, @Aries-ckt, @NuoweiWang, @csunny, @darius-gs, @fangyinc, @fayfox, @jiangshengdev, @polBachelin and @xiuzhu9527
v0.5.0
✨ Enhancements
- Support more chat flows (#1180)
- Add rag awel operator view metadata. (#1174)
- App i18n (#1173)
- Gpts app v0.4 (#1169)
- Native data AI application framework based on AWEL+AGENT (#1152)
🐞 Bug fixes
- Fix api path error (#1165)
🛠️ Other improvements
- Add upgrade to v0.5.0 documents (#1176)
- Update description and dbgpts (#1162)
- Update charts display (#1149)
Thank you to all our contributors for making this release possible!
@Aralhi, @Aries-ckt, @csunny, @fangyinc, @Hzh_97, @lcxadml, @xuyuan23 and @yhjun1026
v0.4.7
✨ Enhancements
- Add support for Azure openAI (#1137)
- Add Jina Embeddings (#1105)
- Refactor for core SDK (#1092)
- Feat/sync docs (#1088)
- Add schema-linking awel example (#1081)
🐞 Bug fixes
- Fix MySQL Database not support DDL init and upgrade. (#1133)
- Fix default setting prompt with history messages (#1117)
- Fix examples/awel get model_name from model_config (#1112)
- Fix bug of sharing data across DAGs (#1102)
- Fix examples/awel default loading model text2vec-large-chinese issue (#1095)
🛠️ Other improvements
- Add documents for dbgpt core lib (#1129)
- Add contributor wall (#1111)
- Fix package name conflict error (#1099)
Thank you to all our contributors for making this release possible!
@Aries-ckt, @JoanFM, @csunny, @fangyinc, @junewgl, @lcxadml, @likenamehaojie and @xiuzhu9527
v0.4.6
✨ Enhancements
- UI component rendering in agent dialog mode (#1083)
- Support chat agent scene use dbgpts (#1075)
- Add retrieve summary agent for both local files and URLs (#1063)
- Add rag operators and rag awel examples (#1061)
- 🎨 Unified color theme, AntV light/dark theme switching, Antd first-screen style loading. (#1020)
- Add summary assistant checking (#1053)
- APP use new SDK component (#1050)
- Multi agents v0.1 (#1044)
- New MessageConverter and more AWEL operators (#1039)
- Add Summary Assistant Agent According to User's Question (#1023)
🐞 Bug fixes
- Fix summary support zhipu message_converter (#1085)
- Fix clean extra file & fix agents use zhipu (#1079)
- Fix awel embedding examples and delete unuseful code. (#1073)
- Fix fix sql component error in the web. (#1068)
- Fix device MPS load llama error (#1033)
🛠️ Other improvements
- Format agent code (#1077)
- Add pylint for DB-GPT core lib (#1076)
- Update wechat QR code (#1069)
- Refactor proxy LLM (#1064)
- Update wechat QRcode (#1046)
- Update chat_knowledge.md and wechat.jpg (#1035)
Thank you to all our contributors for making this release possible!
@2976151305, @Aries-ckt, @csunny, @fangyinc, @qidanrui and @yhjun1026