What's Changed
- Updated the comment in llm_core by @Luckfort in #465
- update apis for launching LSFS terminal by @dongyuanjushi in #466
- add patch for fixing LSFS apis by @dongyuanjushi in #467
- initial commit by @WujiangXu in #468
- update temperature and max new tokens by @dongyuanjushi in #469
- add refer for vllm tool calling by @dongyuanjushi in #470
- update uv installation guide in README by @dongyuanjushi in #472
- docs: update README by @evison in #473
- docs: update readme and publication list by @evison in #475
- fix llm tool call api and update README by @dongyuanjushi in #474
- docs: update the supported models and references in readme by @evison in #477
- fix: hf model tool calling and test scripts by @dongyuanjushi in #476
- fix agent submit proxy by @dongyuanjushi in #482
- add test script and debug storage module by @lin-shh in #484
- refactor launch script to load from config by @dongyuanjushi in #485
- fix tool call decoding for sglang as backend by @dongyuanjushi in #486
- feat. Add Novita LLM API provider as the backend. by @viktor2077 in #487
- fix: sglang tool calling response format by @dongyuanjushi in #488
- fix: configuration refresh by @dongyuanjushi in #490
- refactor llm routing and update tests for routers by @dongyuanjushi in #491
- fix tool call decoding of litellm completion response by @dongyuanjushi in #494
- update model selection logic by @dongyuanjushi in #495
- fix duplicate github action after merging by @dongyuanjushi in #492
- update model selection by @dongyuanjushi in #497
- feat: add environment with computer-use MCP by @dongyuanjushi in #500
- feat(rust): add initial aios-rs scaffold with core traits and README … by @rajakarthik-zenteiq-ai in #505
- Arthur Update README.md by @DheekshithDev in #507
- feat: Qdrant vector database support by @Anush008 in #506
- fix: Docker-ignore
__pycache__in subdirectories by @tomokinakamaru in #509 - fix: Update AIOS core components to fix the issue of tool calling under weak LLMs by @evison in #510
- fix: Add new model version pull for ollama by @evison in #511
- Fix: handle empty response_message after storage operation by @younghosck in #516
New Contributors
- @lin-shh made their first contribution in #484
- @viktor2077 made their first contribution in #487
- @rajakarthik-zenteiq-ai made their first contribution in #505
- @DheekshithDev made their first contribution in #507
- @Anush008 made their first contribution in #506
- @tomokinakamaru made their first contribution in #509
- @younghosck made their first contribution in #516
Full Changelog: v0.2.2...v0.3.0