Try to reuse existing chunks (#3983)

Add support to iframe chatbot (#3929)

Fix errors detected by Ruff (#3918)

Add pagerank to KB. (#3809)

Fix the bug that the agent could not find the context (#3795)

Add api for sessions and add max_tokens for tenant_llm (#3472)

Introduced beartype (#3460)

Move settings initialization after module init phase (#3438)

Use consistent log file names, introduced initLogger (#3403)

Rework logging (#3358)

Integration with Infinity (#2894)

Fix:#3230 When parsing a docx file using the Book parsing method, to_page is always -1, resulting in a block count of 0 even if parsing is successful (#3249)

Fix some issues in API (#2902)

refine API token application (#2847)

Common: Support postgreSQL database as the metadata db. (#2357)

Add a default value for do_refer in Dialog (#2383)

add retry count to task (#2152)

add support for TTS model (#2095)

Format file format from Windows/dos to Unix (#1949)

add support for novita.ai (#1910)

Add agent api (#1888)

trival (#1760)

Fix index=true (#1734)

fix MiniMax api error (#1567)

refine db connection (#1494)

support graph (#1152)

add rerank model (#969)

Update db_models.py

Set the language default value of the language based on the LANG envi… (#853)

The default max tokens of 215 is too small, answers are often cut off.I will modify it to 512 to address this issue. (#845)

let file in knowledgebases visible in file manager (#714)

add file management feature (#560)

Refactor (#537)

make sure the models will not be load twice (#422)

conversation API backend update (#360)

refine log format (#312)

Update default language to English (#309)

change defualt language to English (#185)

apply pep8 formalize (#155)

refine page ranges (#147)

refine manual parser (#140)

add Moonshot, debug my_llm (#126)

add dockerfile for cuda envirement. Refine table search strategy, (#123)

add local llm implementation (#119)

refine pdf parser, add time zone to userinfo (#112)

fix position extraction bug (#93)

fix user login issue (#85)

init README of deepdoc, add picture processer. (#71)

rename vision, add layour and tsr recognizer (#70)

use onnx models, new deepdoc (#68)