github mostlygeek/llama-swap v200

12 hours ago

Changelog

  • d87f0ce docker/unified: publish rootless image variant (#630)
  • 06bc6a6 proxy: preserve wall-clock duration in metrics (#629)
  • a37b486 proxy: add configurable HTTP timeouts for models and peers (#619)
  • 981910d ci: validate config.example.yaml against config-schema.json (#627)
  • a185efe docker: make CMAKE_CUDA_ARCHITECTURES configurable via build arg (#625)
  • 1dd1aad docker/unified: add ik_llama.cpp to CUDA container (#620)
  • 9559009 add /sdapi to list of supported endpoints
  • c2c8cfa docker/unified: build llama.cpp with static libraries (#616)
  • 1e44077 ci: fix matrix exclude for scheduled docker workflow (#610)
  • c794273 docker/unified,.github: fix unified build (#606)
  • 6574a52 build(deps): bump picomatch from 4.0.3 to 4.0.4 in /ui-svelte (#605)

Don't miss a new llama-swap release

NewReleases is sending notifications on new releases.