skulk-llama-server-vulkan
The pip-installable Vulkan llama-server build for Skulk's served GGUF engine (Linux x86_64).
The wheel carries the Foxlight-built llama-server and ggml-rpc-server binaries, compiled from the pinned upstream llama.cpp release with the Vulkan backend, plus the Khronos Vulkan loader (Apache-2.0) bundled alongside, so it has no Python dependencies at all. The system's one prerequisite is the GPU driver's Vulkan ICD: mesa-vulkan-drivers on AMD, the NVIDIA driver's ICD on bare-metal NVIDIA.
uv pip install skulk-llama-server-vulkan
llama-server-vulkan --list-devices
Skulk's engine provisioning discovers the installed wheel automatically and wires it as the node's served engine. Version scheme: 0.<llama.cpp build>.<packaging revision>, in lockstep with the CUDA sibling wheel and Skulk's engine pin. Built from source in the Skulk repository's engine-wheel workflow and published via PyPI trusted publishing.
This wheel is the engine path for AMD GPU nodes and the fallback for bare-metal NVIDIA nodes when the CUDA wheel is unavailable; how it fits into Skulk's install and provisioning flow is documented in the Build & Runtime Paths guide.
The bundled binaries derive from llama.cpp (MIT) and the bundled libvulkan is the Khronos Vulkan Loader (Apache-2.0); both license texts ship in the wheel under skulk_llama_server_vulkan/licenses/.
1 maintainer
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distributions
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file skulk_llama_server_vulkan-0.10092.0-py3-none-manylinux_2_35_x86_64.whl.
File metadata
- Download URL: skulk_llama_server_vulkan-0.10092.0-py3-none-manylinux_2_35_x86_64.whl
- Upload date:
- Size: 65.3 MB
- Tags: Python 3, manylinux: glibc 2.35+ x86-64
- Uploaded using Trusted Publishing? Yes
- Uploaded via: twine/6.1.0 CPython/3.13.14
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
2ba819f697de4198bb8a9bc402d4ef47429ae444f8a1ff14fac0ab30bfdd7d8c
|
|
| MD5 |
b4f9af1d41fb81aa0ffb3ba418617254
|
|
| BLAKE2b-256 |
5fb71a746c55ce54df8b1789bace51eee37617f886333107f70eddf4269e327e
|
Provenance
The following attestation bundles were made for skulk_llama_server_vulkan-0.10092.0-py3-none-manylinux_2_35_x86_64.whl:
Publisher:
engine-wheel.yml on Foxlight-Foundation/Skulk
-
Statement:
-
Statement type:
https://in-toto.io/Statement/v1 -
Predicate type:
https://docs.pypi.org/attestations/publish/v1 -
Subject name:
skulk_llama_server_vulkan-0.10092.0-py3-none-manylinux_2_35_x86_64.whl -
Subject digest:
2ba819f697de4198bb8a9bc402d4ef47429ae444f8a1ff14fac0ab30bfdd7d8c - Sigstore transparency entry: 2226103430
- Sigstore integration time:
-
Permalink:
Foxlight-Foundation/Skulk@872d52bea98b76142762d8edb8abf362f1467e87 -
Branch / Tag:
refs/heads/dev - Owner: https://github.com/Foxlight-Foundation
-
Access:
public
-
Token Issuer:
https://token.actions.githubusercontent.com -
Runner Environment:
github-hosted -
Publication workflow:
engine-wheel.yml@872d52bea98b76142762d8edb8abf362f1467e87 -
Trigger Event:
workflow_dispatch
-
Statement type:
