Commit Graph

263 Commits

Author SHA1 Message Date
wangdan-fit2cloud 689e74af4b perf: 更新slogan文案
Some checks are pending
sync2gitee / repo-sync (push) Waiting to run
Typos Check / Spell Check with Typos (push) Waiting to run
2024-09-10 10:39:22 +08:00
shaohuzhang1 f9a76d7948
feat: 支持openai接口 #353 (#1128) 2024-09-09 14:47:25 +08:00
CaptainB 70f44b990c refactor: 格式规范的pdf通过目录来分段 2024-09-06 10:56:27 +08:00
CaptainB 57b15a8a7f feat: 知识库支持上传csv和excel
--story=1016154 --user=刘瑞斌 【知识库】-支持上传表格类型文档(Excel/CSV)按行分段 https://www.tapd.cn/57709429/s/1567910
2024-08-30 15:46:20 +08:00
zhangshaohu cfb6307293 fix: 修复新增段落迁移文档未索引
Some checks are pending
sync2gitee / repo-sync (push) Waiting to run
Typos Check / Spell Check with Typos (push) Waiting to run
2024-08-28 00:30:18 +08:00
shaohuzhang1 a9443a638c fix: 修复上传文档中后缀为PDF 不识别 2024-08-27 14:16:03 +08:00
shaohuzhang1 b40b4e5305 fix: 修改swagger文档初始化
Some checks are pending
sync2gitee / repo-sync (push) Waiting to run
Typos Check / Spell Check with Typos (push) Waiting to run
2024-08-27 11:44:31 +08:00
shaohuzhang1 df172b530c fix: 修复上传PDF文件智能分段时提示 分段内容不能超过102400个字符 #998 2024-08-26 14:17:25 +08:00
shaohuzhang1 bb6f5d6096 fix: 修复应用保存报错 2024-08-26 13:38:38 +08:00
shaohuzhang1 9e0ac81f1d refactor: 重构模型参数代码 2024-08-23 17:47:23 +08:00
wxg0103 228f913d9b fix: 修复知识库手动添加分段报错的缺陷
--bug=1045628 --user=王孝刚 【知识库】手动添加分段报错 https://www.tapd.cn/57709429/s/1568695
2024-08-23 16:58:02 +08:00
shaohuzhang1 bee832994b fix: 修复函数库输入参数校验,搜索函数失败,编辑函数描述为空校验 2024-08-22 11:50:37 +08:00
shaohuzhang1 d315c01133 fix: 修复dev 无法启动celery 2024-08-22 10:49:22 +08:00
CaptainB 2a87af6172 chore: 解析错误时输出错误原因 2024-08-22 10:43:48 +08:00
shaohuzhang1 ec4fe833b1 fix: 修复无法启动问题
Some checks are pending
sync2gitee / repo-sync (push) Waiting to run
Typos Check / Spell Check with Typos (push) Waiting to run
2024-08-21 18:39:18 +08:00
shaohuzhang1 63e7e52a35 fix: 解决local_model服务端口冲突问题 2024-08-21 16:26:56 +08:00
zhangshaohu 7c5957e0a3 feat: 分离任务 2024-08-21 14:46:38 +08:00
shaohuzhang1 00af530d27
chore: 解析错误时输出错误原因 (#996)
Some checks are pending
sync2gitee / repo-sync (push) Waiting to run
Typos Check / Spell Check with Typos (push) Waiting to run
Co-authored-by: CaptainB <bin@fit2cloud.com>
2024-08-20 22:03:58 +08:00
CaptainB 17af603397 refactor: 优化pdf加载,修复部分pdf中文乱码的问题
Some checks are pending
sync2gitee / repo-sync (push) Waiting to run
Typos Check / Spell Check with Typos (push) Waiting to run
2024-08-20 16:58:04 +08:00
CaptainB 01d8204cb5 refactor: 逐页加载pdf, 图片类型单独保存成文件加载 2024-08-16 15:08:22 +08:00
CaptainB 0d59ab2be9 refactor: 使用lazy_load方式加载pdf
Some checks are pending
sync2gitee / repo-sync (push) Waiting to run
Typos Check / Spell Check with Typos (push) Waiting to run
2024-08-16 10:43:20 +08:00
CaptainB e266dd9d99 refactor: 支持解析pdf中的图片
Some checks are pending
sync2gitee / repo-sync (push) Waiting to run
Typos Check / Spell Check with Typos (push) Waiting to run
2024-08-15 20:53:44 +08:00
shaohuzhang1 f421d1975d feat: 函数库功能 2024-08-15 17:17:25 +08:00
shaohuzhang1 0ad5a76598
fix: 修复分段时,特殊情况会丢失数据 #938 (#946)
Some checks are pending
sync2gitee / repo-sync (push) Waiting to run
Typos Check / Spell Check with Typos (push) Waiting to run
2024-08-07 19:52:05 +08:00
shaohuzhang1 864bca6450
fix: 修复索引中的文档,知识库删除后依然再执行 (#934) 2024-08-06 16:22:53 +08:00
shaohuzhang1 b3c7120372
fix: 修复QA文件解析失败 (#933) 2024-08-06 14:47:28 +08:00
shaohuzhang1 d33bf6a8e8
fix: 修复知识库问题管理修改问题报错 (#929) 2024-08-06 10:13:52 +08:00
shaohuzhang1 76c1acbabb
fix: 修复讯飞星火模型多用户同时提问后,回答内容会错乱 #917 (#920) 2024-08-02 10:39:54 +08:00
shaohuzhang1 16d7316dca
feat: 段落分块设置最小分块字数 (#898) 2024-07-30 11:08:53 +08:00
shaohuzhang1 6aa2a682bc
fix: 修复向量化失败,缓存未被删除 (#887) 2024-07-29 16:17:30 +08:00
shaohuzhang1 3c7142ed7c
feat: 文档添加排队中状态 (#886) 2024-07-29 15:51:33 +08:00
shaohuzhang1 d935e9a836
fix: 修复上传文档,高级分段设置分段长度为10w字符,生成预览还是4096个字符一段 (#884) 2024-07-29 14:08:40 +08:00
shaohuzhang1 1979cf12c4
feat: gunicorn启动使用gthread (#882)
Some checks failed
sync2gitee / repo-sync (push) Has been cancelled
Typos Check / Spell Check with Typos (push) Has been cancelled
2024-07-26 18:58:01 +08:00
shaohuzhang1 c76bc0f2c2
refactor: 重构部分代码 (#865) 2024-07-25 11:54:41 +08:00
shaohuzhang1 0d2524df1d
refactor: 重构部分代码 (#864) 2024-07-25 10:41:38 +08:00
shaohuzhang1 0131f46e37
fix: 修改启动方式 (#860) 2024-07-24 16:55:00 +08:00
shaohuzhang1 53434f9d24
feat: 细分段落chunk增加召回命中率 (#841) 2024-07-23 18:19:41 +08:00
shaohuzhang1 5f3f1dd2ca
feat: 显示设置功能 2024-07-22 18:52:56 +08:00
shaohuzhang1 c465ddff19
fix: 修复无法流式输出 (#820)
Some checks failed
sync2gitee / repo-sync (push) Has been cancelled
Typos Check / Spell Check with Typos (push) Has been cancelled
2024-07-19 21:34:17 +08:00
shaohuzhang1 16851592c3
feat: 添加web服务器gunicorn 2024-07-19 18:23:56 +08:00
gcalgoz 4d8ac28674
fix: 优化文档分割处理 (#814) 2024-07-19 17:33:18 +08:00
shaohuzhang1 a9b8bdd365 feat: 支持向量模型 2024-07-19 16:44:53 +08:00
shaohuzhang1 b14a799350 feat: 支持向量模型 2024-07-19 10:34:47 +08:00
wangdan-fit2cloud 57aade8d62 feat 外观设置 2024-07-19 10:15:24 +08:00
shaohuzhang1 9e76cd97de feat: 支持向量模型 2024-07-18 16:36:34 +08:00
shaohuzhang1 dcf5892b96 feat: 支持向量模型 2024-07-18 15:44:48 +08:00
shaohuzhang1 75b9b17e2e feat: 支持向量模型 2024-07-18 10:26:16 +08:00
shaohuzhang1 bd4303aee7 feat: 支持向量模型 2024-07-17 17:01:57 +08:00
shaohuzhang1 fd6aa4fb39
fix: 优化认证逻辑 (#724)
Some checks are pending
sync2gitee / repo-sync (push) Waiting to run
Typos Check / Spell Check with Typos (push) Waiting to run
2024-07-09 18:59:15 +08:00
shaohuzhang1 60e65a8b17
feat: 接口校验 (#721)
feat: 接口校验
2024-07-09 13:44:27 +08:00
shaohuzhang1 c1e1a19a42
feat: 添加文件上传接口 (#712)
Some checks are pending
sync2gitee / repo-sync (push) Waiting to run
Typos Check / Spell Check with Typos (push) Waiting to run
2024-07-05 19:02:20 +08:00
gcalgoz d152f441ff
fix: 对模型嵌入归一化,提升检索速度 (#710) 2024-07-05 18:14:58 +08:00
wangdan-fit2cloud b902bf153e feat: 规范字体 2024-07-05 15:32:13 +08:00
shaohuzhang1 00f2a8bbd4
fix: 修复【知识库】高级分段 自动清洗,把所有换行被去除 (#684) 2024-07-02 11:32:20 +08:00
RatishT 6c67b65e6a
Fix pylint warning `no-else-return` (#678) 2024-07-02 10:55:26 +08:00
shaohuzhang1 2f2f74fdab
feat: 支持工作流 (#671) 2024-07-01 09:45:59 +08:00
Evan 726ba2a8e2
fix:知识库上传文档操作报异常,错误发生在 ts_vecto_util.py 文件的 replace_word 函数中:错误的原因是正则表达式模式在某个位置缺少一个右括号,使得子模式未完全结束。该提交修复了这个问题。 (#667)
Some checks are pending
sync2gitee / repo-sync (push) Waiting to run
Typos Check / Spell Check with Typos (push) Waiting to run
2024-06-27 09:51:14 +08:00
shaohuzhang1 7a5bfa2673
fix: 修复导入带有表格样式的md文件分段后表格格式丢失 #615 (#630)
Some checks failed
sync2gitee / repo-sync (push) Has been cancelled
Typos Check / Spell Check with Typos (push) Has been cancelled
2024-06-13 10:41:52 +08:00
evilstar 3a7c4d3568
修复:关键词提取的bug (#621)
Some checks failed
sync2gitee / repo-sync (push) Has been cancelled
Typos Check / Spell Check with Typos (push) Has been cancelled
2024-06-11 18:25:12 +08:00
shaohuzhang1 cdc7e13415
feat: 升级langchain版本到2.2.3 (#617)
Some checks failed
sync2gitee / repo-sync (push) Has been cancelled
Typos Check / Spell Check with Typos (push) Has been cancelled
2024-06-07 11:08:32 +08:00
shaohuzhang1 11a8df151e
fix: 修复通过API进行对话的API文档接口错误 #587 (#593) 2024-06-03 10:53:03 +08:00
shaohuzhang1 ed7ddfbc59
fix: 修复分段超过分段长度限制 (#577) 2024-05-29 12:00:56 +08:00
shaohuzhang1 dd22fa0868
perf: 优化向量化文档后修改更新时间 (#571) 2024-05-28 18:12:45 +08:00
shaohuzhang1 22e192ed11
fix: 修复文档导入解析错误 (#570)
Some checks are pending
sync2gitee / repo-sync (push) Waiting to run
Typos Check / Spell Check with Typos (push) Waiting to run
2024-05-28 17:32:29 +08:00
shaohuzhang1 efe5a2b021
fix: 修复excel导入失败问题 (#554) 2024-05-27 16:06:14 +08:00
shaohuzhang1 e9a05b1255
fix: 修复qa知识库导入失败错误 (#536) 2024-05-24 17:59:02 +08:00
shaohuzhang1 7953706895
perf: 优化错误提示 (#533) 2024-05-24 16:18:33 +08:00
shaohuzhang1 2da31997e5
perf: 优化API_KEY调用对话文档 (#532) 2024-05-24 15:57:43 +08:00
shaohuzhang1 a3af104ef0
feat: 知识库增加重新向量化功能 2024-05-24 11:27:59 +08:00
shaohuzhang1 28938104c0
* feat: 支持上传 Excel/CSV 类型的问答对 (#430)
Some checks are pending
sync2gitee / repo-sync (push) Waiting to run
Typos Check / Spell Check with Typos (push) Waiting to run
2024-05-23 18:57:49 +08:00
shaohuzhang1 86f500208f
feat: 支持上传html格式的文档 #364 (#518) 2024-05-23 14:19:18 +08:00
shaohuzhang1 84c7728d00
fix: 修复应用修改模型后,历史对话未使用最新模型 (#487)
Some checks are pending
sync2gitee / repo-sync (push) Waiting to run
Typos Check / Spell Check with Typos (push) Waiting to run
2024-05-20 20:21:45 +08:00
shaohuzhang1 7f30d03abd
fix: 修复分词超过数据库最大限制 (#401) 2024-05-09 15:55:35 +08:00
shaohuzhang1 3fb6192021
fix: 跨域失效 (#394) 2024-05-08 18:46:58 +08:00
shaohuzhang1 267be441e3
feat: 跨域设置(#276) 2024-05-08 17:13:13 +08:00
shaohuzhang1 d4e742f7c6
feat: 分段管理支持批量迁移,删除分段 #113,#103 2024-05-08 10:40:15 +08:00
shaohuzhang1 c1b6ec630c
fix: 导入文档中含有特殊字符时,导入失败。 #363 (#372) 2024-05-07 12:18:03 +08:00
shaohuzhang1 a788d8f3b8
perf: 优化超长文本rsa加密解密 (#312) 2024-04-29 13:53:49 +08:00
shaohuzhang1 cd472b47f7
fix: 模型添加长字符的加密解密方式 (#310) 2024-04-29 13:28:47 +08:00
shaohuzhang1 29427a0ad6
fix: 私有部署计算tokens报错 (#284) 2024-04-28 11:59:19 +08:00
shaohuzhang1 9d808b4ccd
feat: 支持文档迁移(#52) 2024-04-26 18:35:46 +08:00
shaohuzhang1 1f916a5c3e
feat: 【知识库】docx支持图片上传 #69 (#267) 2024-04-26 18:03:02 +08:00
shaohuzhang1 84e8ca0f84
fix: 应用修改获取对话详情被拒绝 (#251) 2024-04-25 16:17:29 +08:00
shaohuzhang1 b26265fefd
feat: 【应用】支持自定义上传应用的logo #54
* feat: 【知识库】本地上传的文档内带的图片能同步到 maxkb 里 #69 
* feat: 【应用】支持自定义上传应用的logo #54
2024-04-23 19:03:34 +08:00
shaohuzhang1 df54fd2b54
fix: 对话API调用,输出报错 #192 (#200) 2024-04-22 13:52:46 +08:00
shaohuzhang1 c89ae29429
feat: 增加全文检索和混合检索方式 2024-04-22 11:21:24 +08:00
feng626 11f0a82e68
perf: Optimize the writing method of getting HTTP_AUTHORIZATION (#114) 2024-04-16 22:55:34 +08:00
shaohuzhang1 5e499e6afa
fix: PDF上传知识库开始导入接口报错 #122 (#125) 2024-04-16 20:59:27 +08:00
shaohuzhang1 3d28d2527b
fix: typos
* fix: typos
2024-04-15 19:06:42 +08:00
Eric_Lee dfb04b1052
perf: 优化代码和错误提示 (#101) 2024-04-15 16:53:26 +08:00
taojinlong 944ba01661 fix: 统一为中文简体 2024-04-15 15:43:37 +08:00
wxg0103 765ae8c79f
docs: 修改注释的错别字 (#78) 2024-04-15 14:51:09 +08:00
shaohuzhang1 260060f276 fix: 优化同步网页逻辑 2024-04-10 18:43:43 +08:00
shaohuzhang1 8b31fd6b36 fix: 分段不支持类型的文件报错 2024-04-10 17:05:46 +08:00
shaohuzhang1 bd3f6e4a9b fix: word分段支持表格数据 2024-04-10 10:38:17 +08:00
shaohuzhang1 765c79ed9d fix: 修改分段正则,优化分段逻辑 2024-04-09 18:05:50 +08:00
shaohuzhang1 b038f12a52 fix: 上传文档大小扩大到100MB 2024-04-09 15:21:40 +08:00
shaohuzhang1 11d8c6f174
fix: 修改已知bug(#30)
* fix: 刷新公共访问链接后,客户端统计重置

* fix: 导出未提交的sql文件

* fix: 创建 MaxKB 在线文档的知识库,只能获取根地址数据,子地址数据无法获取
2024-04-02 19:32:04 +08:00
shaohuzhang1 16ab1f0eae
Pr@main@fix bugs (#27)
* fix: 优化word分段规则

* fix: 去除标题特殊字符

* fix: 对话重新生成问题

---------

Co-authored-by: wangdan-fit2cloud <dan.wang@fit2cloud.com>
2024-04-01 14:39:56 +08:00
shaohuzhang1 c55bb3f6e5
Pr@main@pdf (#23)
* feat: 分段API支持word,pdf

* fix: 通用型知识库支持上传 PDF/DOC 格式的文档#19

---------

Co-authored-by: wangdan-fit2cloud <dan.wang@fit2cloud.com>
2024-03-29 18:28:05 +08:00
shaohuzhang1 05d01144f9 fix: 修改统计对话接口,添加时间段统计 2024-03-28 16:06:54 +08:00
shaohuzhang1 cf003aa2d2 fix: 同步web站点内容编码错误,导致乱码 2024-03-25 18:46:25 +08:00
shaohuzhang1 d732a46f89 fix: 【知识库】导入非utf8 编码的txt文件,分段内容是空白 2024-03-25 11:05:15 +08:00
zhangshaohu 53d45e069d fix: 智能分段文本分段丢失数据 2024-03-23 19:46:20 +08:00
shaohuzhang1 2d91a0f2bf fix: 程序启动将正在下载的模型设置为失败状态 2024-03-22 20:18:59 +08:00
shaohuzhang1 d074424398 feat: ollama支持下载模型 2024-03-22 17:56:56 +08:00
shaohuzhang1 d59aac4a2a fix: 删除用户报错 2024-03-21 18:33:35 +08:00
shaohuzhang1 28d44ac567 feat: 去除无用代码 2024-03-21 14:49:21 +08:00
shaohuzhang1 40031bd29d feat: 密钥对存储到数据库 2024-03-20 11:57:17 +08:00
shaohuzhang1 f7b9677a8c feat: 表迁移 2024-03-18 17:54:48 +08:00
shaohuzhang1 4e93c6f4c7 feat: 用户管理相关接口 2024-03-18 15:34:02 +08:00
shaohuzhang1 27d8285388 fix: 静态资源白名单,添加非空判断 2024-03-14 14:30:50 +08:00
shaohuzhang1 e39c813c15 feat: 每日凌晨重置客户访问数量 2024-03-14 12:26:42 +08:00
zhangshaohu 0fbd5873f7 feat: 客户端不使用cookie存储改为localstore,优化认证代码 2024-03-14 05:43:01 +08:00
shaohuzhang1 21a557ef43 fix: 白名单逻辑修改 2024-03-13 21:50:57 +08:00
shaohuzhang1 fa36e6bbab feat: 添加嵌入访问限制,白名单 2024-03-13 16:07:13 +08:00
shaohuzhang1 b470b1b6e5 feat: 添加问题管理相关接口,兼容历史版本 2024-03-11 17:28:05 +08:00
shaohuzhang1 d266b83d84 fix: 导入文档标题不能超过256个字符,修改统一响应异常 2024-03-06 18:37:11 +08:00
shaohuzhang1 ba3e4e7556 feat: 对接ollama平台模型 2024-03-06 13:43:45 +08:00
shaohuzhang1 da3015fa36 fix: 分段正则修改 2024-03-04 18:34:47 +08:00
shaohuzhang1 b153ca9e59 fix: 【知识库】知识库设置,web站点地址格式错误,保存报错 2024-03-04 11:01:58 +08:00
shaohuzhang1 f7a8f11ef7 fix: 校验提示 2024-03-04 10:12:18 +08:00
shaohuzhang1 855684472d fix: 嵌入认证跨域问题 2024-03-01 14:25:56 +08:00
shaohuzhang1 30a19a0ed2 fix: 【知识库】知识库使用tagname选择器,有部分页面没有导入数据 2024-03-01 11:14:32 +08:00
shaohuzhang1 cc62c35995 fix: 分段时 title超过256字符将超出部分拼接给content 2024-02-29 18:48:10 +08:00
shaohuzhang1 911f00737a fix: 【权限】有应用管理权限,启用停用apikey、删除apikey提示没有权限 2024-02-29 16:14:07 +08:00
shaohuzhang1 22c319a2bf fix: 同步知识库,无法获取内容 2024-02-29 15:14:53 +08:00
shaohuzhang1 8450b3598c fix: 同步web站点知识库 解析md 未按照标签解析 2024-02-29 12:02:29 +08:00
shaohuzhang1 7eb18fbf30 fix: 上传的文档中 未智能处理空白段落 2024-02-19 11:29:24 +08:00
shaohuzhang1 04f34d748e fix: 【知识库】整体同步,只删除了没有同步 2024-01-29 17:07:07 +08:00
shaohuzhang1 68f3515e83 fix: 获取网络文档去掉ssl校验 2024-01-25 15:25:07 +08:00
shaohuzhang1 ef66af639e fix: 同步web站点 选择器不填无法获取内容 2024-01-24 11:23:16 +08:00
shaohuzhang1 67e6138066 fix: 批量删除文档,未删除关联段落信息, 添加关联问题报错 2024-01-23 15:52:15 +08:00
shaohuzhang1 344c336143 fix: web站点 批量同步,批量删除,批量导入接口 2024-01-17 16:08:51 +08:00
shaohuzhang1 3f87335c80 feat: 优化对话逻辑 2024-01-16 16:46:54 +08:00
shaohuzhang1 82c8e322fb feat: web站点数据集文档同步 2024-01-03 11:51:48 +08:00
shaohuzhang1 edbc8561c7 fix: 分段错误,会话模板修改 2024-01-02 15:34:19 +08:00
shaohuzhang1 64c8cc6b39 feat: web数据集 2023-12-29 18:02:23 +08:00
shaohuzhang1 8fb268c25a feat: url获取文档数据工具 2023-12-27 18:33:23 +08:00
shaohuzhang1 a7704c3a8a feat: 命中率测试接口 2023-12-25 17:10:59 +08:00
shaohuzhang1 ce1c2271eb fix: 分段换行被替换 2023-12-21 17:46:43 +08:00
shaohuzhang1 2c6135a929 feat: 问答时,同步存入日志,优化向量化执行逻辑,修改model下载目录 2023-12-21 16:55:11 +08:00
shaohuzhang1 b6f7537c2b feat: 日志打印,嵌入脚本 2023-12-21 12:16:39 +08:00
shaohuzhang1 9750a550a5 feat: 创建数据集向量化,以文档的粒度向量化 2023-12-20 15:23:59 +08:00
shaohuzhang1 81df321782 fix: 邮件格式 2023-12-20 14:57:45 +08:00
shaohuzhang1 b61da14fdd fix: 去除分段日志打印 2023-12-20 14:39:35 +08:00
shaohuzhang1 f6d84b5c53 fix: 修改向量化段落拼接 2023-12-20 13:56:25 +08:00
shaohuzhang1 2273081d52 fix: 修改日志答案保存到段落后,关联问题未做向量化处理。 2023-12-20 12:36:09 +08:00
wangdan-fit2cloud a82d282255 feat: 2023-12-18 11:32:29 +08:00
shaohuzhang1 740f1d3dd1 fix: 向量化的时候限制最大值 2023-12-15 14:22:19 +08:00