Pull requests

• fix(ui): Offer multi-GPU device selections in recipe settings

  #3999·fl0rianr opened Oct 28, 2026Collaborator·8 of 12 tasks·

  ·
  3

• fix(server): remove global inference timeout and decouple operationl deadline

  #3998·fl0rianr opened Oct 28, 2026Collaborator·8 of 12 tasks·

  ·
  5

• chore: bump pwilkin backend versions

  ci:backends
  #3993·pwilkin opened Oct 28, 2026Collaborator·8 of 12 tasks·

  ·
  4

• fix(auto-tune): correct AMD UMA budgets and preserve CPU offload

  area::apicppengine::llamacppruntime::rocm
  #3988·fl0rianr opened Oct 27, 2026Collaborator·10 of 15 tasks·

  ·
  7

• feat(backends): Add support for Strata v0.2.7

  #3987·julianxhokaxhiu opened Oct 27, 2026Contributor·7 of 12 tasks·

  ·
  11

• fix(server): guard chat message content extraction against JSON null

  bugcppengine::vllm
  #3983·mike2307 opened Oct 27, 2026Contributor·8 of 11 tasks·

  ·
  1

  1

• fix(ci): validate sd-cpp auto-updates and flag failed checks

  area::cicppengine::sd
  #3975·fl0rianr opened Oct 26, 2026Collaborator·8 of 12 tasks·

  ·
  5

• fix(app): support Firefox chatbot sidebar prompts

  area::api
  #3973·fl0rianr opened Oct 25, 2026Collaborator·8 of 12 tasks·

  ·
  17

• Update stable-diffusion.cpp to master-1006-90e86d1 and CUDA master-1006-90e86d1

  #3970·github-actions[bot] opened Oct 25, 2026Bot·

  ·
  1

• fix(ci): make llama.cpp auto-update documentation failures visible

  area::cibugci:backendsengine::llamacpprfc:not-required
  #3965·fl0rianr opened Oct 24, 2026Collaborator·8 of 12 tasks·

  ·
  6

• Declare Vulkan loader/driver as weak deps for deb/rpm packages; document Ubuntu Server prerequisites

  #3961·superm1 with CopilotAI opened Oct 23, 2026·7 of 12 tasks·

  ·
  1

  3

  

• fix(router): preserve pinned models during load recovery

  area::apibugcpp
  #3960·mikemikimike opened Oct 23, 2026·

  ·
  1

  1

• Fixes the GUI3 banner in the news page

  area::api
  #3959·meghsat opened Oct 23, 2026Collaborator·7 of 12 tasks·

  ·
  2

• Add SystemOne decision models: /v1/systemone, GGUF zero-shot on /v1/classify, and the systemone router classifier

  area::apiengine::llamacpp
  #3956·meghsat opened Oct 23, 2026Collaborator·5 of 12 tasks·

  ·
  3

• Add AMICA backend support to Lemonade

  area::apicppenhancement
  #3953·Zhaeong opened Oct 23, 2026·

  ·
  12

• Adding zero-shot classifier support

  area::apicppenhancement
  #3952·meghsat opened Oct 23, 2026Collaborator·8 of 11 tasks·

  ·
  29

• Feature/ryzenai vision detections

  area::apiengine::ryzenai
  #3949·ali-haidous opened Oct 22, 2026·12 tasks·

  ·
  1

• fix(server): harden chat and Anthropic response handling

  rfc:not-required
  #3937·fl0rianr opened Oct 22, 2026Collaborator·7 of 11 tasks·

  ·
  1

  1

• fix(trellis): correct HTTP error handling and watchdog retrial

  ci:backendsengine::llamacpp
  #3935·fl0rianr opened Oct 22, 2026Collaborator·8 of 12 tasks·

  ·
  1

• Integrate Nexus SDK, device management, and mesh inference selection

  ci:distrosci:macos
  #3930·Geramy opened Oct 22, 2026Member·

  ·
  45

• feat(containers): run container backends from lemond.service with rootless Podman

  area::installerci:backendscppenhancement
  #3911·jeremyfowers opened Oct 19, 2026Member·2/6·4 of 8 tasks·

  ·
  1

  

• fix(gui3): remove stub vendor mermaid 11

  area::installer
  #3908·fl0rianr opened Oct 19, 2026Collaborator·7 of 12 tasks·

  ·
  6

• feat: rework server code to use JSON configs instead of binary wrappers

  area::apiengine::llamacpp
  #3906·pwilkin opened Oct 19, 2026Collaborator·

  ·
  2

• Update llama.cpp to b11719

  #3905·github-actions[bot] opened Oct 18, 2026Bot·

  ·
  1

• fix(router): skip the evict-and-retry when no other model is loaded

  area::api
  #3894·bernardogv opened Oct 17, 2026Contributor·8 of 12 tasks·

  ·
  1