Release v0.4.3 · langgenius/dify

Optimize performance & fix few bugs.

What's Changed

fix: zhipuai chatglm turbo prompts must user, assistant in sequence by @takatost in #1899
fix: customize model schema fetch failed raise error by @takatost in #1900
Fix/add qdrant timeout default value by @JohnJyong in #1901
fix: remove useless code by @zxhlyh in #1903
fix: read file encoding error by @newsouther in #1902
fix: tongyi generate error by @takatost in #1904
fix: array oob in azure openai embeddings by @crazywoola in #1905
fix: spark credentials validate by @takatost in #1906
fix: notion_indexing_estimate embedding_model_instance NPE by @takatost in #1907
fix: window size changed causes result regeneration by @iamjoel in #1908
fix: model quota format by @zxhlyh in #1909
fix: remove useless code. by @GarfieldDai in #1913
ci: Properly cache pip packages by @bowenliang123 in #1912
web: Add style CI workflow to enforce eslint checks on web module by @bowenliang123 in #1910
fix customer spliter character by @JohnJyong in #1915
feat: model parameter prefefined by @zxhlyh in #1917
fix: email template style by @takatost in #1914
Fix comparison bug in ApplicationQueueManager by @waltcow in #1919
fix: app create raise error when no available model providers by @takatost in #1921
feat: optimize invoke errors by @takatost in #1922
alter schedule timedelta by @JohnJyong in #1923
fix a typo in file agent_app_runner.py by @liupeng0913 in #1927
fix: xinference cache by @Yeuoly in #1926
feat: optimize performance by @takatost in #1928
feat: bump version 0.4.3 by @takatost in #1930

Full Changelog: 0.4.2...0.4.3