sci-ml/ik_llama-cpp (bentoo)

Search

Install

Install this package:

emerge -a sci-ml/ik_llama-cpp

Package Information

Description:
ik_llama.cpp is a fork of llama.cpp by Iwan Kawrakow, the author of most of the k-quants, carrying additional state-of-the-art quantization types (the IQK families) and CPU matrix-multiplication kernels that are substantially faster than upstream for quantized inference. It builds the same llama-cli, llama-server, llama-quantize and llama-bench programs as sci-ml/llama-cpp. To let both be installed at once, every binary is installed with an ik_ prefix (ik_llama-cli, ik_llama-server, and so on) and the shared libraries live in a private subdirectory of libdir, so nothing collides with sci-ml/llama-cpp. No headers or pkg-config files are installed: this package is an application, not a development target. Link against sci-ml/llama-cpp instead. Upstream publishes no release tags, so the package is pinned to a dated commit snapshot fetched as a commit archive. The CPU backend is tuned through CPU_FLAGS_X86. This matters more here than for sci-ml/llama-cpp, because the fork's quantization kernels are written for AVX2 and AVX-512; without those flags the generic fallback gives up most of the fork's advantage. Keywording is deliberately restricted to ~amd64. The SOTA quants are hand-tuned for AVX2/AVX-512, the ARM paths are far less exercised upstream, and no arm64 build of this snapshot has been verified. Note that this ebuild is adapted from the guru overlay (sci-misc/ik_llama-cpp).
Homepage:
https://github.com/ikawrakow/ik_llama.cpp
License:
MIT

Versions

Version EAPI Keywords Slot
0_pre20260802 8 ~amd64 0

Metadata

Description

Maintainers

Upstream

Raw Metadata XML
<pkgmetadata>
	<maintainer type="person">
		<email>lucascs@proton.me</email>
		<name>Lucas C.S.</name>
	</maintainer>
	<longdescription lang="en">
		ik_llama.cpp is a fork of llama.cpp by Iwan Kawrakow, the author of most
		of the k-quants, carrying additional state-of-the-art quantization types
		(the IQK families) and CPU matrix-multiplication kernels that are
		substantially faster than upstream for quantized inference.

		It builds the same llama-cli, llama-server, llama-quantize and
		llama-bench programs as sci-ml/llama-cpp. To let both be installed at
		once, every binary is installed with an ik_ prefix (ik_llama-cli,
		ik_llama-server, and so on) and the shared libraries live in a private
		subdirectory of libdir, so nothing collides with sci-ml/llama-cpp. No
		headers or pkg-config files are installed: this package is an
		application, not a development target. Link against sci-ml/llama-cpp
		instead.

		Upstream publishes no release tags, so the package is pinned to a dated
		commit snapshot fetched as a commit archive.

		The CPU backend is tuned through CPU_FLAGS_X86. This matters more here
		than for sci-ml/llama-cpp, because the fork's quantization kernels are
		written for AVX2 and AVX-512; without those flags the generic fallback
		gives up most of the fork's advantage.

		Keywording is deliberately restricted to ~amd64. The SOTA quants are
		hand-tuned for AVX2/AVX-512, the ARM paths are far less exercised
		upstream, and no arm64 build of this snapshot has been verified.
		Note that this ebuild is adapted from the guru overlay
		(sci-misc/ik_llama-cpp).
	</longdescription>
	<use>
		<flag name="hip">Build a HIP (ROCm) backend</flag>
	</use>
	<upstream>
		<bugs-to>https://github.com/ikawrakow/ik_llama.cpp/issues</bugs-to>
		<remote-id type="github">ikawrakow/ik_llama.cpp</remote-id>
	</upstream>
</pkgmetadata>

Lint Warnings

USE Flags

Manage flags for this package: euse -i <flag> -p sci-ml/ik_llama-cpp | euse -E <flag> -p sci-ml/ik_llama-cpp | euse -D <flag> -p sci-ml/ik_llama-cpp

Flag Description 0_pre20260802
"( ⚠️
( ⚠️
) ⚠️
)" ⚠️
avx ⚠️
avx2 ⚠️
avx512_bf16 ⚠️
avx512_vnni ⚠️
avx512bw ⚠️
avx512cd ⚠️
avx512dq ⚠️
avx512f ⚠️
avx512vbmi ⚠️
avx512vl ⚠️
avx_vnni ⚠️
cuda Build cycles renderer with nVidia CUDA support. ⚠️
curl Enable support for automated uploading of crash reports ⚠️
f16c ⚠️
fma3 ⚠️
hip Build a HIP (ROCm) backend
openmp ⚠️
vulkan Add support for the Vulkan viewport backend ⚠️

Manifest

Type File Size Versions
DIST ik_llama-cpp-0_pre20260802.tar.gz 34738658 bytes 0_pre20260802
Unmatched Entries
Type File Size