v0.5.0-27

janhq/cortex.cpp

版本发布时间: 2024-08-05 20:54:20

janhq/cortex.cpp最新发布版本:v0.5.1-rc1(2024-09-12 14:31:21)

Changes

chore: return error on pulling an existing model @louis-jan (#975)
feat: add config cpp port @marknguyen1302 (#973)
chore: clean logs periodically @louis-jan (#972)
Fix codesign macos for application and installer @hiento09 (#974)
fix: persists engine version on init @louis-jan (#969)
feat: support multi process in cortex @marknguyen1302 (#965)
Update the Base URL for the API reference to reflect the new URL @irfanpena (#971)
fix: disable devtool @marknguyen1302 (#967)
Fix codesign windows @hiento09 (#964)
fix: create message request with string content instead of array @louis-jan (#963)
fix: cortex engines init api with empty body should init default options @louis-jan (#962)
fix: change default host cortex @marknguyen1302 (#961)
feat: watch models and engines update for proper data retrieval @louis-jan (#960)
fix: fix issue when re-download aborted model @marknguyen1302 (#959)
Add launchpad uninstaller prerm and postrm @hiento09 (#957)
feat: support martian nvidia engine @marknguyen1302 (#956)
chore: delete local file when abort download @marknguyen1302 (#954)
fix: terminate system should kill cortex process @louis-jan (#955)
fix: start and run models are not outputting last error logs @louis-jan (#951)
fix: remove hardcode stream @marknguyen1302 (#952)
chore: persist context length and ngl from gguf file @louis-jan (#947)
chore: show error from remote engine, lint @marknguyen1302 (#949)
Fix openai api pipeline @hiento09 (#948)
chore: destroy dangling processes on uninstall @louis-jan (#945)
feat: support openrouter, cohere engine @marknguyen1302 (#946)
feature: support local model pull @louis-jan (#944)
chore: handle failed download @marknguyen1302 (#943)
chore: specify engine version to pull @louis-jan (#942)
bump cortex llamacpp for lla‌ma 3.1 @Van-QA (#941)