71 Revīzijas (617847e3c0b9aad0b1e0eca40198c453c6b652e0)

Autors SHA1 Ziņojums Datums
  sino d27e3ab99d
chore: remove unresolved reference (#6110) pirms 1 gada
  Bowen Liang dcb72e0067
chore: apply flake8-comprehensions Ruff rules to improve collection comprehensions (#5652) pirms 1 gada
  Jyong ba5f8afaa8
Feat/firecrawl data source (#5232) pirms 1 gada
  Bowen Liang f976740b57
improve: mordernizing validation by migrating pydantic from 1.x to 2.x (#4592) pirms 1 gada
  takatost d1dbbc1e33
feat: backend model load balancing support (#4927) pirms 1 gada
  Jyong b6631cd878
modify rerank and splitter code directory (#4924) pirms 1 gada
  Jyong 233c4150d1
support images and tables extract from docx (#4619) pirms 1 gada
  Rain Chen c255a20d7c
allow to config max segmentation tokens length for RAG document using environment variable (#4375) pirms 1 gada
  Bowen Liang 04ad46dd31
chore: skip unnecessary key checks prior to accessing a dictionary (#4497) pirms 1 gada
  LIU HONGWEI c227f3d985
feat: Deprecate datetime.utcnow() in favor of datetime.now(timezone.utc).replace(tzinfo=None) for better timezone handling (#3408) (#3416) pirms 1 gada
  Jyong 33ea689861
fix detached instance error in keyword index create thread and fix question classifier node out of index error (#3219) pirms 1 gada
  Jyong 283979fc46
fix keyword index error when storage source is S3 (#3182) pirms 1 gada
  takatost 7753ba2d37
FEAT: NEW WORKFLOW ENGINE (#3160) pirms 1 gada
  Jyong b0b0cc045f
add mutil-thread document embedding (#3016) pirms 1 gada
  Jyong 6454e1d644
chunk-overlap None check (#2781) pirms 1 gada
  Jyong 31070ffbca
fix qa index processor tenant id is None error (#2713) pirms 1 gada
  Charlie.Wei fa7ba30ba3
Fix rebuild index&csv parsing (#2705) pirms 1 gada
  Jyong 5b953c1ef2
Fix some RAG bugs (#2570) pirms 1 gada
  Jyong 0620fa3094
Feat/vdb migrate command (#2562) pirms 1 gada
  Jyong 4be3087642
Fix/new RAG bugs (#2547) pirms 1 gada
  Jyong 91ea6fe4ee
Fix/langchain document schema (#2539) pirms 1 gada
  Jyong 6c4e6bf1d6
Feat/dify rag (#2528) pirms 1 gada
  Jyong 97fe817186
Fix/upload limit (#2521) pirms 1 gada
  Bowen Liang 063191889d
chore: apply ruff's pyupgrade linter rules to modernize Python code with targeted version (#2419) pirms 1 gada
  crazywoola 243ca5b1e2
fix: typo in package path of core.splitter (#2411) pirms 1 gada
  Bowen Liang 843280f82b
enhancement: introduce Ruff for Python linter for reordering and removing unused imports with automated pre-commit and sytle check (#2366) pirms 1 gada
  takatost 9f637ead38
bump version to 0.5.3 (#2306) pirms 1 gada
  KVOJJJin 89fcf4ea7c
Feat: chunk overlap supported (#2209) pirms 1 gada
  takatost 6cf93379b3
fix: split chunks return empty strings (#2197) pirms 1 gada
  Jyong 869690c485
fix notion estimate (#2090) pirms 1 gada
  Jyong cb7a608d75
ascii filter Unicode U+FFFE (#2038) pirms 1 gada
  Jyong a63a9c7d45
text spliter length method use default embedding model tokenizer (#2011) pirms 1 gada
  Bowen Liang cc9e74123c
improve: introduce isort for linting Python imports (#1983) pirms 1 gada
  Jyong 24bdedf802
fix get embedding model provider in empty dataset (#1986) pirms 1 gada
  Jyong 4a3d15b6de
fix customer spliter character (#1915) pirms 1 gada
  takatost a938e1f184
fix: notion_indexing_estimate embedding_model_instance NPE (#1907) pirms 1 gada
  Yeuoly 9134849744
fix: remove tiktoken from text splitter (#1876) pirms 1 gada
  takatost d069c668f8
Model Runtime (#1858) pirms 1 gada
  Jyong df1509983c
ppt & pptx improve (#1790) pirms 1 gada
  Jyong 5e34f938c1
Feat/add unstructured support (#1780) pirms 1 gada
  crazywoola 994fceece3
fix: qa regex (#1738) pirms 1 gada
  Pascal M bc54cdc537
refactor: typo in dataset docstore (#1711) pirms 1 gada
  Pascal M 5d10cf0fe6
fix: error Class 'builtins.list' is not mapped (#1710) pirms 1 gada
  Jyong 4588831bff
Feat/add retriever rerank (#1560) pirms 1 gada
  crazywoola d0e1ea8f06
1506 remove duplicated code (#1511) pirms 2 gadiem
  Garfield Dai 42a5b3ec17
feat: advanced prompt backend (#1301) pirms 2 gadiem
  Jyong 289c93d081
Feat/improve document delete logic (#1325) pirms 2 gadiem
  yezhwi 8b8e510bfe
fix: handle AttributeError for datasets and index (#1052) pirms 2 gadiem
  Jyong a55ba6e614
Fix/ignore economy dataset (#1043) pirms 2 gadiem
  Jyong 2d604d9330
Fix/filter empty segment (#1004) pirms 2 gadiem