Skip to content

0.4.11

Compare
Choose a tag to compare
@github-actions github-actions released this 15 Apr 07:50
26ce5eb

Changes (#2627)

🚀 Features

  • feat: nitro additional dependencies @louis-jan (#2674)
  • feat: move log into monitoring extension @louis-jan (#2662)
  • Feature playwright integrate report portal @hiento09 (#2648)

🐛 Fixes

  • fix: add new shortcut to create new thread @louis-jan (#2701)
  • fix: broken thread settings - component disabled @louis-jan (#2696)
  • fix: allow smaller min width @louis-jan (#2695)
  • fix: add markdown support for extension description @namchuai (#2691)
  • fix: not able to update model setting in local api @namchuai (#2687)
  • fix: race condition issue - reading settings.json file @louis-jan (#2683)
  • fix: better kill process tensorrt-llm @louis-jan (#2681)
  • fix: wrong monitoring system information type @louis-jan (#2679)
  • fix: can't read the setting at first time @namchuai (#2677)
  • fix: add fallback as default endpoint for inference engine @namchuai (#2669)
  • fix: api server chat completion error for remote model @namchuai (#2671)
  • Change report portal description to git commit hash @hiento09 (#2668)
  • fix: missing api key error handling @louis-jan (#2645)
  • fix: unload model while loading cause unknown error @louis-jan (#2649)
  • fix: api server - server url prefix is not updated @louis-jan (#2650)
  • fix: model selector lack of warning message when no models downloaded @louis-jan (#2651)
  • fix: redirect to extension setting @louis-jan (#2626)
  • fix: quick ask could not register global shortcuts @louis-jan (#2642)
  • fix: object is destroyed block app exit @louis-jan (#2623)
  • fix: cancel loading model with stop action @louis-jan (#2607)
  • fix: copied logs line break @louis-jan (#2603)
  • fix: duplicated error message @louis-jan (#2602)
  • fix: wrong response header for non streamed responses @louis-jan (#2606)
  • fix: app does not gracefully stop a model @louis-jan (#2593)
  • fix: does not register global shortcut when quick ask is disabled @louis-jan (#2592)
  • fix: groq frequently stops during long responses @louis-jan (#2584)
  • fix: regenerate message should not include the assistant's previous response @louis-jan (#2583)
  • fix: missing load model error message @louis-jan (#2581)
  • fix: broken auto link @louis-jan (#2580)
  • fix: disabling Vulkan causes users to be stuck in GPU-accelerated mode @louis-jan (#2578)
  • fix: quick ask blocks app update @louis-jan (#2525)

🧰 Maintenance

  • chore: Revert to 0.3.16 due to Nitro issue @Van-QA (#2700)
  • chore: Revert bump to 0.3.14 due to Nitro issue @Van-QA (#2699)
  • bump nitro version to 0.3.21 @hiento09 (#2680)
  • chore: Extension should have product name in manifest @louis-jan (#2675)
  • Enable ci linter workflow_dispatch event @hiento09 (#2673)
  • Change report portal description to git commit hash @hiento09 (#2668)
  • Bump nitro to 0.3.19 @hiento09 (#2663)
  • Bump nitro version to 0.3.18 @hiento09 (#2652)
  • chore: filter gguf files when importing model on windows @namchuai (#2653)
  • chore: add GPU driver and toolkit status @louis-jan (#2628)
  • Correct condition checking cuda dependencies windows @hiento09 (#2629)
  • chore: extension should register its own models @namchuai (#2601)
  • Change condition for trigger PR from forked repo @hiento09 (#2611)
  • docs: Update README.md @0xSage (#2612)
  • Enable trigger for pull_request_target ci linter @hiento09 (#2609)
  • chore: add open log dir to troubleshooting modal @louis-jan (#2605)
  • chore: update place holder for api key of groq and openai @namchuai (#2588)
  • Move out from cloudflare r2 to aws s3 @hiento09 (#2596)
  • Change release download url to cloudflare worker proxy and update download model tensorrt llm to aws s3 endpoint @hiento09 (#2576)
  • Change npm registry to nexus for CI test and enable turbo remote cache @hiento09 (#2535)
  • chore: some wordings in extension settings @namchuai (#2573)

Contributor

@0xSage, @GenkaOk, @Inchoker, @Van-QA, @hahuyhoang411, @hiento09, @jan-service-account, @louis-jan, @myakura, @namchuai and @zhhanging