gks

One package for working with GGUF models locally: an OpenAI-compatible LLM server, a diffusion image/video/audio generator and a GGUF metadata/tensor editor with a built-in quantizer — three panels on one GUI, powered by one unified engine compiled in a single build on top of gk, an independent tensor library. There is no ggml anywhere in the tree.

Install

pip install gks

The build compiles the bundled engine (CPU by default, Metal on macOS). GPU backends are opt-in at install time:

GKS_CUDA=1 pip install gks     # NVIDIA
GKS_HIP=1 pip install gks      # AMD ROCm
GKS_VULKAN=1 pip install gks   # Vulkan

Each switch drives the whole engine — the server, the diffusion runtime and the multimodal projectors all evaluate their graphs on the one gk build.

Run

gks                 # unified GUI — Server / Diffuser / Editor panels
python -m gks       # same thing

Each panel also runs on its own, exactly like the standalone gguf-server / gguf-diffusion / gguf-editor packages did:

gks server          # LLM server GUI
gks diffuser        # image generation GUI
gks editor          # GGUF editor GUI

And the engines are directly scriptable from the CLI:

gks server engine -- --model model.gguf --port 8888
gks diffuser engine -- -m sd.gguf -p "a lighthouse at dusk" -o out.png
gks editor quantize -m in.gguf -o out-q4_k.gguf --type q4_k
gks editor devices

Layout

vendor/engine/           the unified gks engine (one CMake build)
  gk/                    the gk compute kernels (CPU + optional GPU backends)
  gk/compat/             the historical ggml C API, implemented on gk
  src/ common/ mtmd/     GGUF LLM runtime
  app/                   the gguf-server HTTP server
  diffusion/             diffusion runtime + CLI
  quantizer/             quantizer shared library (its own quant kernels)
src/gks/                 the Python package
  server/ diffuser/ editor/   the three panels (backend + web frontend each)
  gui.py static/         the unified 3-panel GUI shell

Nothing above gk/compat/ knows gk exists: the runtimes include the same ggml.h / ggml-backend.h / gguf.h headers and call the same functions they always did, while graph building, allocation, scheduling and the kernels themselves are gk's. See vendor/engine/README.md for the engine's own build options.

The editor's quantizer stays independent — its qz_* codec is compiled both into the quantizer library the editor drives and into gk itself, so the encoder and the runtimes' decoder can never disagree about a GGUF block.