-
Notifications
You must be signed in to change notification settings - Fork 476
All issues
Issue creation is restricted in this repository
Issues
is:issue state:open
is:issue state:open
Search results
- Status: Open.
Windows status bar: report VRAM usage and GPU utilization
enhancementNew feature or requestNew feature or requestStatus: Open.#3317 In lemonade-sdk/lemonade;acestep vulkan backend: bundled libggml-cpu.so requires AVX2 — SIGILL (exit 132) on pre-AVX2 x86-64 CPUs
bugSomething isn't workingSomething isn't workingengine::flmFastFlowLM backend (NPU); multi-modal LLM/ASR/embeddings/rerankingFastFlowLM backend (NPU); multi-modal LLM/ASR/embeddings/rerankingruntime::cpuCPU-only execution pathCPU-only execution pathStatus: Open.#3302 In lemonade-sdk/lemonade;Model appears in downloaded as two different variants, though only downloaded one, deleting one deletes the other
bugSomething isn't workingSomething isn't workingenhancementNew feature or requestNew feature or requestStatus: Open.#3298 In lemonade-sdk/lemonade;Models moved from "default lemonade model directory" TO "external custom models directory" don't display correctly
bugSomething isn't workingSomething isn't workingStatus: Open.#3297 In lemonade-sdk/lemonade;GUI3: model download quirk
bugSomething isn't workingSomething isn't workingStatus: Open.GUI3: MTP models not downloading correctly from Hugging Face, MTP file missing
bugSomething isn't workingSomething isn't workingStatus: Open.Specification: Integration of Specialized llama.cpp Engine Forks (CachyLLama and ROCmFPX) into Lemonade Server
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 requestStatus: Open.#3279 In lemonade-sdk/lemonade;GUI3: weird behaviour with the "Additional backend CLI arguments"
bugSomething isn't workingSomething isn't workingStatus: Open.#3278 In lemonade-sdk/lemonade;llama.cpp: --ctx-size is applied per slot, so concurrent requests exhaust the KV cache
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.#3276 In lemonade-sdk/lemonade;MCP: expose model management capabilities used by GUI3
area::apiHTTP REST API surface and route handlersHTTP REST API surface and route handlersStatus: Open.#3275 In lemonade-sdk/lemonade;GUI3: discover Lemonade capabilities through MCP tools/list
area::apiHTTP REST API surface and route handlersHTTP REST API surface and route handlersenhancementNew feature or requestNew feature or requestStatus: Open.#3274 In lemonade-sdk/lemonade;