-
Notifications
You must be signed in to change notification settings - Fork 452
All issues
Issue creation is restricted in this repository
Issues
is:issue state:open
is:issue state:open
Search results
Feature Request: model anti-affinity (per-model and per-tier) to prevent co-resident OOM
enhancementNew feature or requestNew feature or requestStatus: Open.#2919 In lemonade-sdk/lemonade;lemond serves IPv6-only when the IPv4 bind fails, and reports itself healthy
bugSomething isn't workingSomething isn't workingStatus: Open.#2917 In lemonade-sdk/lemonade;Bump sdcpp versions to >= 727
engine::sdstable-diffusion.cpp backend; image generation/edit/variationsstable-diffusion.cpp backend; image generation/edit/variationsenhancementNew feature or requestNew feature or requestStatus: Open.#2912 In lemonade-sdk/lemonade;Videocard selection
engine::llamacppllama.cpp backend (LlamaCppServer); GPU/CPU LLM inference (Vulkan, ROCm, Metal)llama.cpp backend (LlamaCppServer); GPU/CPU LLM inference (Vulkan, ROCm, Metal)enhancementNew feature or requestNew feature or requestruntime::rocmAMD ROCm runtimeAMD ROCm runtimeStatus: Open.#2890 In lemonade-sdk/lemonade;Update the maintainers table in contribute.md based on recent activity
documentationImprovements or additions to documentationImprovements or additions to documentationStatus: Open.#2888 In lemonade-sdk/lemonade;feat: Benchmark reporting platform
area::ciCI / GitHub Actions / self-hosted runner infrastructureCI / GitHub Actions / self-hosted runner infrastructureenhancementNew feature or requestNew feature or requesttype:featureNew capabilityNew capabilityStatus: Open.Unable to run any models on llama.cpp:cuda in Docker - 500 errors
bugSomething isn't workingSomething isn't workingengine::llamacppllama.cpp backend (LlamaCppServer); GPU/CPU LLM inference (Vulkan, ROCm, Metal)llama.cpp backend (LlamaCppServer); GPU/CPU LLM inference (Vulkan, ROCm, Metal)runtime::cudaNVIDIA CUDA runtimeNVIDIA CUDA runtimeStatus: Open.#2879 In lemonade-sdk/lemonade;llamacpp.rocm_bin: latest serves a stale binary (de69995) despite installer logging an upgrade to the newest build
bugSomething isn't workingSomething isn't workingengine::llamacppllama.cpp backend (LlamaCppServer); GPU/CPU LLM inference (Vulkan, ROCm, Metal)llama.cpp backend (LlamaCppServer); GPU/CPU LLM inference (Vulkan, ROCm, Metal)runtime::rocmAMD ROCm runtimeAMD ROCm runtimeStatus: Open.#2875 In lemonade-sdk/lemonade;Installing backend:llamacpp:cuda uses TAR on 7zip requires Windows 11 22H2+ breaking Win10 and older Win11 support
area::installerWindows MSI / macOS DMG / Debian / RPM packagingWindows MSI / macOS DMG / Debian / RPM packagingbugSomething isn't workingSomething isn't workingengine::llamacppllama.cpp backend (LlamaCppServer); GPU/CPU LLM inference (Vulkan, ROCm, Metal)llama.cpp backend (LlamaCppServer); GPU/CPU LLM inference (Vulkan, ROCm, Metal)runtime::cudaNVIDIA CUDA runtimeNVIDIA CUDA runtimeStatus: Open.#2861 In lemonade-sdk/lemonade;Allow pulling a model at a specific Hugging Face revision, so a model can be locked to one commit
area::apiHTTP REST API surface and route handlersHTTP REST API surface and route handlersenhancementNew feature or requestNew feature or requestStatus: Open.#2860 In lemonade-sdk/lemonade;feat: Backends Performance Leaderboard/QA Tracking for performance
enhancementNew feature or requestNew feature or requestStatus: Open.#2858 In lemonade-sdk/lemonade;extract_zip() missing strip-components/wrapper-folder handling — causes silent extraction failures on Windows (e.g. flm:npu)
bugSomething isn't workingSomething isn't workingengine::flmFastFlowLM backend (NPU); multi-modal LLM/ASR/embeddings/rerankingFastFlowLM backend (NPU); multi-modal LLM/ASR/embeddings/rerankingStatus: Open.#2856 In lemonade-sdk/lemonade;