65 Commits

Author	SHA1	Message	Date
ChPi	ab24b1777a	(fix)(common) prevent NullPointerException for jsonFormat (#2365)	2025-08-17 16:21:36 +08:00
jerryjzhang	1f6d217b26	[fix]Fix unit test cases.	2025-08-05 17:53:58 +08:00
ChPi	af28bc7c2a	(improvement)(parser) Add json format to LLM request for performance improvement (#2352)	2025-08-05 17:43:03 +08:00
beat4ocean	c4992501bd	[Improvement][headless] Only expose OPEN_AI/OLLAMA/DIFY chat model providers. (#2237)	2025-05-05 15:51:26 +08:00
beat4ocean	acffc03c79	[Improvement][headless] Replace deprecated LangChain4j APIs (#2235)	2025-05-05 15:47:51 +08:00
jerryjzhang	1fed8ca4d9	(fix)(headless)Fix hash code issue of MetricSchemaResp and DimensionSchemaResp.	2025-03-30 07:20:07 +08:00
beat4ocean	232a202275	[Improvement] Upgrade dependencies and fix vulnerabilities. (#2190)	2025-03-28 09:07:49 +08:00
jerryjzhang	6cc145935d	[fix][auth]Fix user registration and resetPassword issue.	2025-02-14 00:28:36 +08:00
zyclove	d942d35c93	feat:add opensearch (#2049)	2025-02-12 20:30:58 +08:00
zyclove	cb139a54e8	feat:add openapi supports ApiVersion (#2050)	2025-02-12 20:28:10 +08:00
lexluo09	8b69d57c4b	[improvement][chat] Fix the issue with the DatabaseMatchStrategy variable under multi-threading (#1963)	2024-12-19 10:04:17 +08:00
wwsheng009	b57eed47e2	SAP HANA DATABASE Source support improvement[优化SAPhana数据库的支持] (#1959) ... * (improvement)(database) update the support for sap hana database source * (fix)(common) add the default timeout for ZhipuAiEmbeddingModel,avoid the program error	2024-12-17 21:52:19 +08:00
Jun Zhang	e55f43c737	[improvement]Use QueryWrapper in place of hard-coded SQLs (#1944) ... * [improvement][launcher]Use API to get element ID avoiding hard-code. * [fix][launcher]Fix mysql scripts. * [improvement][launcher]Support DuckDB database and refactor translator code structure. * [improvement][headless-fe] Revamped the interaction for semantic modeling routing and successfully implemented the switching between dimension and dataset management. * [improvement][Headless] Add table ddl in Dbschema * [improvement][Headless] Add get database by type * [improvement][Headless] Supports automatic batch creation of models based on db table names. * [improvement][Headless] Supports getting domain by bizName * [improvement][launcher]Refactor unit tests and demo data. * [fix][launcher]Change default vector dimension to 512. * [improvement](Dict) add dimValueAliasMap info for KnowledgeBaseService * [improvement][headless]Use QueryWrapper to replace hard-code SQL in mapper xml. * [improvement][chat]Introduce ChatMemory to delegate ChatMemoryDO. * [fix][common]Fix embedding store sys configs. * [fix][common]Fix postgres schema, using varchar instead of char. * [improvement][launcher]Change supersonic docker deployment from mysql to postgres. * [Fix][launcher]Fix a number of issues related to semantic modeling. * [Fix][headless]Fix the evaluation logic of agg type. * [fix][assembly]Fix Dockerfile and add docker compose run script. * [fix][chat]Fix "multiple assignments to same column "similar_queries". * [improvement][headless]Use LamdaQueryWrapper to avoid hard-coded column names. * [improvement][headless]Refactor headless infra to support advanced semantic modelling. * [improvement][headless]Change class name `Dim` to `Dimension`. * [improvement][chat]Introduce `TimeFieldMapper` to always map time field. * [fix][headless]Remove unnecessary dimension existence check. * [fix][chat]Fix adjusted filters don't take effect. ---------	2024-12-08 13:32:29 +08:00
lexluo09	3a905d7fb1	[improvement][chat] Milvus supports setting the user and password parameters (#1845)	2024-10-24 22:59:35 +08:00
zhaodongsheng	0ddcdf93ec	(feat)Support for managing large models with Dify #1830;2、add user access token; #1829; 3、support change password #1824 (#1839)	2024-10-22 13:58:58 +08:00
lexluo09	d7f301064a	[improvement][common] Fix the issue of missing 'ORDER BY' clause generation (#1802)	2024-10-15 10:22:22 +08:00
Zhengyang Jia	1ef642d0dd	[improvement][common]Add pgvector vector library adaptation. (#1800)	2024-10-15 09:16:32 +08:00
lexluo09	2c75a6972a	(improvement)[build] Format the enumeration class with line breaks (#1761)	2024-10-09 20:56:08 +08:00
jerryjzhang	cd57d2344e	[fix][chat]Model provider is case sensitive.#1739	2024-10-09 17:48:42 +08:00
jerryjzhang	248f4f83f6	[feature][chat]Refactor chat model config related codes.#1739	2024-10-09 17:27:07 +08:00
lexluo09	71a9954be5	(improvement)[build] Use Spotless to customize the code formatting (#1750)	2024-10-04 00:05:04 +08:00
lexluo09	4b1dab8e4a	(improvement)(headless) Upgrade to the latest version of langchain4j and add support for embedding deletion operation and reset. (#1660)	2024-09-12 18:16:16 +08:00
lexluo09	879696e493	[improvement](chat) Optimize the Schema Mapping rules (#1649)	2024-09-10 22:58:34 +08:00
daikon	ffc84dfc4c	(improvement)(common) update OpenAiModelFactory default model (#1641)	2024-09-09 12:27:46 +08:00
lexluo09	5f59e89eea	(improvement)(build) Add spotless during the build process. (#1639)	2024-09-07 00:36:17 +08:00
lexluo09	afa82bf98d	(improvement)(chat) Support connecting to Milvus service in the cloud. (#1588)	2024-08-21 16:43:51 +08:00
lexluo09	10a5e485cb	(improvement)(chat) Add unit tests for each chatModel and embeddingModel. (#1582)	2024-08-18 23:43:47 +08:00
lexluo09	115cf19078	(improvement)(chat) Modify the default URL and model of zhipu, and enable search support for qwen. (#1578)	2024-08-17 23:12:48 +08:00
lexluo09	898c7100ba	(improvement)(chat) Modify the configuration of the zhipu vector database to disable the default persistence of the vector database configuration. (#1577)	2024-08-16 23:14:23 +08:00
lexluo09	7150f19def	(improvement)(chat) Vector retrieval supports filtering by modelId collection during query. (#1576)	2024-08-16 21:31:07 +08:00
lexluo09	8d29e89317	(improvement)(chat) Fix the path issue in Dashscope for large models and optimize the EmbeddingService code. (#1566)	2024-08-13 13:37:24 +08:00
lexluo09	8b01dac8d4	(improvement)(chat) Support large models qianfan, zhipu, Azure, LocalAi, Dashscope, and handle the apiKey configuration as hidden. (#1552)	2024-08-11 23:28:24 +08:00
lexluo09	9a1fac5d4c	(improvement)(chat) Reduce frequent loading of embedding models to improve loading performance. (#1478)	2024-07-30 11:25:03 +08:00
JackLiang	8f620480c6	fix：加载模型到内存时内存泄露bug (#1465) ... * fix existSegment bug which embedding search not has min-score,so that it will return true.	2024-07-26 18:07:49 +08:00
lexluo09	d64ed02df9	(improvement)(chat) Remove langchain4j configuration file and perform all configuration for the large model through the UI interface. (#1442)	2024-07-20 21:30:46 +08:00
lexluo09	529251097b	(improvement)(chat) Remove initial default values, make the configuration file settings take effect, and optimize the code. (#1406)	2024-07-14 23:20:06 +08:00
lexluo09	d5c78d87e7	(improvement)(chat) Support integrating embeddingStore into system settings. (#1405)	2024-07-14 21:50:12 +08:00
lexluo09	4eb6193699	(improvement)(chat) Integrate llm configuration into system settings. (#1403)	2024-07-14 14:47:17 +08:00
jerryjzhang	5bf4a4160d	(fix)(chat&headless)Agent references ChatModelConfig instead of ModelConfig.	2024-07-12 14:15:54 +08:00
lexluo09	4d7bfe07aa	(improvement)(chat) Support embedding store configuration. (#1363)	2024-07-07 00:30:19 +08:00
lexluo09	6db6aaf98d	(improvement)(chat) Support configuring embeddingModel or embeddingStore at the agent level. (#1361)	2024-07-06 20:44:23 +08:00
lexluo09	93ea7a618c	(improvement)(chat) Fix embedded embedding model, some machines throw an error with version GLIBC_2.27 (#1346)	2024-07-04 21:02:08 +08:00
lexluo09	7773442fbf	(improvement)(chat) Fix the error in Milvus query and add the option to create EmbeddingStore based on caching mode (#1310)	2024-07-01 16:29:43 +08:00
二毛	7bb0f84bc3	fix milvus shutdown error (#1307)	2024-07-01 14:55:40 +08:00
lexluo09	da43f790c3	(improvement)(project) Remove unnecessary logs and remove redundant utility classes in the Dockerfile (#1306)	2024-07-01 14:42:30 +08:00
lexluo09	c8b6af4887	(improvement)(chat) It supports successful startup using Docker, local, and other methods. If the Chroma store creation fails, it will be replaced with in-memory (#1301)	2024-06-30 20:15:46 +08:00
lexluo09	528491717b	(improvement)(headless) Add support for the Ollama provider in the frontend and optimize the code (#1270)	2024-06-28 17:29:58 +08:00
lexluo09	391c0dccc8	(improvement)(headless) Add deduplication and persistence to InMemoryEmbeddingStore (#1256)	2024-06-27 22:24:49 +08:00
二毛	20697efac1	(improvement)(headless) Add localAi auto-configuration. (#1243)	2024-06-27 15:09:18 +08:00
lexluo09	7c711f6105	(improvement)(project) Remove additional logging dependencies and use logback as the primary logging framework. (#1235)	2024-06-27 09:39:26 +08:00