sci-misc/llama-cpp (pf4public)

Search

Install

Install this package:

emerge -a sci-misc/llama-cpp

Package Information

Description:
LLM inference in C/C++
Homepage:
https://github.com/ggml-org/llama.cpp
License:
MIT

Versions

Version EAPI Keywords Slot
9999 8 ~amd64 ~arm64 0
0.8.0 8 amd64 ~arm64 0
0.7.0 8 ~amd64 ~arm64 0

Metadata

Maintainers

Upstream

Raw Metadata XML
<pkgmetadata>
	<maintainer type="person">
		<email>PF4Public@users.noreply.github.com</email>
	</maintainer>
	<use>
		<flag name="build-online">Allow the build to download the npm registry (live ebuild only, needed by the webui flag)</flag>
		<flag name="cuda">Enable the NVIDIA CUDA GPU backend</flag>
		<flag name="examples">Build the example programs</flag>
		<flag name="hip">Enable the AMD ROCm/HIP GPU backend</flag>
		<flag name="openblas">Enable the OpenBLAS backend</flag>
		<flag name="opencl">Enable the OpenCL backend</flag>
		<flag name="openmp">Use OpenMP in the CPU backend</flag>
		<flag name="openssl">Use OpenSSL for HTTPS model downloads</flag>
		<flag name="rpc">Enable the RPC backend for distributed inference</flag>
		<flag name="system-ggml">Link against the installed sci-ml/ggml instead of the bundled copy; the CPU ISA set and the backends are then selected by ggml's own USE flags</flag>
		<flag name="vulkan">Enable the Vulkan backend</flag>
		<flag name="webui">Embed the prebuilt web UI in llama-server</flag>
	</use>
	<upstream>
		<bugs-to>https://github.com/ggml-org/llama.cpp/issues</bugs-to>
		<remote-id type="github">ggml-org/llama.cpp</remote-id>
	</upstream>
</pkgmetadata>

Lint Warnings

USE Flags

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

Flag Description 9999 0.8.0 0.7.0
"( ⚠️ ✓ ✓ ✓
( ⚠️ ✓ ✓ ✓
) ⚠️ ✓ ✓ ✓
)" ⚠️ ✓ ✓ ✓
amx_bf16 ⚠️ ✓ ✓ ✓
amx_int8 ⚠️ ✓ ✓ ✓
amx_tile ⚠️ ✓ ✓ ✓
avx ⚠️ ✓ ✓ ✓
avx2 ⚠️ ✓ ✓ ✓
avx512_bf16 ⚠️ ✓ ✓ ✓
avx512_vnni ⚠️ ✓ ✓ ✓
avx512f ⚠️ ✓ ✓ ✓
avx512vbmi ⚠️ ✓ ✓ ✓
avx_vnni ⚠️ ✓ ✓ ✓
bmi2 ⚠️ ✓ ✓ ✓
build-online Allow the build to download the npm registry (live ebuild only, needed by the webui flag) ✓ ✓ ✓
cuda Enable the NVIDIA CUDA GPU backend ✓ ✓ ✓
examples Build the example programs ✓ ✓ ✓
f16c ⚠️ ✓ ✓ ✓
fma3 ⚠️ ✓ ✓ ✓
hip Enable the AMD ROCm/HIP GPU backend ✓ ✓ ✓
openblas Enable the OpenBLAS backend ✓ ✓ ✓
opencl Enable the OpenCL backend ✓ ✓ ✓
openmp Use OpenMP in the CPU backend ⊕ ⊕ ⊕
openssl Use OpenSSL for HTTPS model downloads ⊕ ⊕ ⊕
rpc Enable the RPC backend for distributed inference ✓ ✓ ✓
sse4_2 ⚠️ ✓ ✓ ✓
system-ggml Link against the installed sci-ml/ggml instead of the bundled copy; the CPU ISA set and the backends are then selected by ggml's own USE flags ✓ ✓ ✓
vulkan Enable the Vulkan backend ✓ ✓ ✓
webui Embed the prebuilt web UI in llama-server ✓ ✓ ✓

Files

Manifest

Type File Size Versions
DIST llama-b11669-ui.tar.gz 3140519 bytes 9999, 0.7.0
DIST llama-b11960-ui.tar.gz 3244974 bytes 0.8.0
DIST llama-cpp-0.7.0.tar.gz 38183379 bytes 0.7.0
DIST llama-cpp-0.8.0.tar.gz 38680037 bytes 0.8.0
Unmatched Entries
Type File Size