похуй
This commit is contained in:
parent
e439700992
commit
0591e57dfd
8 changed files with 503 additions and 0 deletions
23
.gear/check
Executable file
23
.gear/check
Executable file
|
|
@ -0,0 +1,23 @@
|
||||||
|
#!/bin/bash -efu
|
||||||
|
|
||||||
|
RET=0
|
||||||
|
|
||||||
|
f=tools/server/public/bundle.js
|
||||||
|
email=$(git log -1 --pretty=%ae -- "$f")
|
||||||
|
if [[ $email != *@altlinux.org ]]; then
|
||||||
|
echo >&2 "$f is not updated by ALT. Perhaps, run .gear/generate"
|
||||||
|
(set -x; git l -1 -- "$f")
|
||||||
|
RET=1
|
||||||
|
fi
|
||||||
|
|
||||||
|
ts_f=$(git log -1 --date=unix --pretty='%cd' "$f")
|
||||||
|
s=tools/server/webui
|
||||||
|
ts_s=$(git log -1 --date=unix --pretty='%cd' "$s")
|
||||||
|
if (( ts_f < ts_s )); then
|
||||||
|
echo >&2 "$f is not updated after $s update."
|
||||||
|
(set -x; git l -1 -- "$f")
|
||||||
|
(set -x; git l -1 -- "$s")
|
||||||
|
RET=1
|
||||||
|
fi
|
||||||
|
|
||||||
|
exit $RET
|
||||||
60
.gear/gen-manpage
Executable file
60
.gear/gen-manpage
Executable file
|
|
@ -0,0 +1,60 @@
|
||||||
|
#!/bin/bash -e
|
||||||
|
|
||||||
|
#
|
||||||
|
# Wrapper around llama-cli to massage the output to be more palatable to
|
||||||
|
# help2man for generating a manual page.
|
||||||
|
#
|
||||||
|
# Invoke as:
|
||||||
|
# gen-manpage build/bin/llama-cli --name='LLM inference in C/C++'
|
||||||
|
#
|
||||||
|
# This script is in the public domain and may be distributed or modified
|
||||||
|
# without restriction.
|
||||||
|
#
|
||||||
|
|
||||||
|
: ${HELP2MAN:=help2man}
|
||||||
|
|
||||||
|
tmp=$(mktemp -d)
|
||||||
|
trap "rm -rf $tmp" EXIT HUP INT
|
||||||
|
|
||||||
|
cmd="${1?}"
|
||||||
|
shift
|
||||||
|
|
||||||
|
# --version goes to stderr?
|
||||||
|
$cmd --version 2>$tmp/version
|
||||||
|
|
||||||
|
# --help to stdout, some diagnostics to stderr
|
||||||
|
$cmd --help >$tmp/help 2>/dev/null
|
||||||
|
|
||||||
|
# Ensure we got something from both.
|
||||||
|
test -s $tmp/version && test -s $tmp/help
|
||||||
|
|
||||||
|
cmd="${cmd##*/}"
|
||||||
|
|
||||||
|
# Strip out just version.
|
||||||
|
sed -i -n "s/^version: /$cmd /p" $tmp/version
|
||||||
|
|
||||||
|
# Convert ---thing--- to subheadings, and clean up args.
|
||||||
|
sed -i \
|
||||||
|
-e 's/^----- \([A-Za-z].*\) -----/\1:/' \
|
||||||
|
-e 's/^[a-z].*:$/\u&/' \
|
||||||
|
-e '/^[A-Za-z].*:$/s/ [a-z]/\U&/g' \
|
||||||
|
-e 's/^-\(-\?[a-zA-Z0-9][^ ]*,\)\( *\)\(-.*\) /-\1 \3\2 /' \
|
||||||
|
-e 's/^-\(.*\) / -\1 /' \
|
||||||
|
-e '/ *[^ ].*: [^ ]/s/:/: /' \
|
||||||
|
$tmp/help
|
||||||
|
|
||||||
|
cat <<EOT >$tmp/${cmd##*/}
|
||||||
|
#!/bin/bash
|
||||||
|
|
||||||
|
for arg; do
|
||||||
|
case "\$arg" in
|
||||||
|
--version) cat $tmp/version;;
|
||||||
|
--help) cat $tmp/help;;
|
||||||
|
esac
|
||||||
|
done
|
||||||
|
EOT
|
||||||
|
|
||||||
|
chmod a+x "$tmp/$cmd"
|
||||||
|
${HELP2MAN} --no-info "$@" "$tmp/$cmd"
|
||||||
|
|
||||||
|
exit $?
|
||||||
354
.gear/llama-cpp-turboquant.spec
Normal file
354
.gear/llama-cpp-turboquant.spec
Normal file
|
|
@ -0,0 +1,354 @@
|
||||||
|
# SPDX-License-Identifier: GPL-2.0-only
|
||||||
|
%define _unpackaged_files_terminate_build 1
|
||||||
|
%define _stripped_files_terminate_build 1
|
||||||
|
%set_verify_elf_method strict
|
||||||
|
|
||||||
|
%ifarch x86_64
|
||||||
|
%def_with cuda
|
||||||
|
%else
|
||||||
|
%def_without cuda
|
||||||
|
%endif
|
||||||
|
%def_with vulkan
|
||||||
|
|
||||||
|
Name: llama-cpp-turboquant
|
||||||
|
Version: 8659
|
||||||
|
Release: alt1
|
||||||
|
Epoch: 1
|
||||||
|
Summary: LLM inference in C/C++
|
||||||
|
License: MIT
|
||||||
|
Group: Sciences/Computer science
|
||||||
|
# https://ggml.ai/
|
||||||
|
Url: https://github.com/ggerganov/llama.cpp
|
||||||
|
ExcludeArch: %ix86
|
||||||
|
Requires: %name-cpu = %EVR
|
||||||
|
%if_with cuda
|
||||||
|
Requires: %name-cuda = %EVR
|
||||||
|
%filter_from_requires /(libcudart\.so\.12)/d
|
||||||
|
%filter_from_requires /debug64(libcuda\.so\.1)/d
|
||||||
|
%endif
|
||||||
|
%if_with vulkan
|
||||||
|
Requires: %name-vulkan = %EVR
|
||||||
|
%endif
|
||||||
|
|
||||||
|
Source: %name-%version.tar
|
||||||
|
Patch: %name-%version.patch
|
||||||
|
|
||||||
|
BuildRequires(pre): rpm-macros-cmake
|
||||||
|
BuildRequires: cmake
|
||||||
|
BuildRequires: gcc-c++
|
||||||
|
BuildRequires: help2man
|
||||||
|
BuildRequires: libcurl-devel
|
||||||
|
BuildRequires: libgomp-devel
|
||||||
|
BuildRequires: libssl-devel
|
||||||
|
BuildRequires: libstdc++-devel-static
|
||||||
|
%if_with cuda
|
||||||
|
BuildRequires: gcc12-c++
|
||||||
|
BuildRequires: nvidia-cuda-devel-static
|
||||||
|
%endif
|
||||||
|
%if_with vulkan
|
||||||
|
BuildRequires: glslc
|
||||||
|
BuildRequires: libvulkan-devel
|
||||||
|
%endif
|
||||||
|
%{?!_without_check:%{?!_disable_check:
|
||||||
|
BuildRequires: ctest
|
||||||
|
BuildRequires: python3-module-jinja2
|
||||||
|
BuildRequires: tinyllamas-gguf
|
||||||
|
}}
|
||||||
|
|
||||||
|
%description
|
||||||
|
Plain C/C++ implementation (of inference of many LLM models) without
|
||||||
|
dependencies. AVX, AVX2, AVX512, and AMX support for x86 architectures.
|
||||||
|
Mixed F16/F32 precision. 1.5-bit, 2-bit, 3-bit, 4-bit, 5-bit, 6-bit, and
|
||||||
|
8-bit integer quantization for faster inference and reduced memory use.
|
||||||
|
Supports CPU, GPU, and hybrid CPU+GPU inference.
|
||||||
|
|
||||||
|
Supported models:
|
||||||
|
|
||||||
|
LLaMA models, Mistral 7B, Mixtral MoE, Falcon, Chinese LLaMA /
|
||||||
|
Alpaca and Chinese LLaMA-2 / Alpaca-2, Vigogne (French), Koala,
|
||||||
|
Baichuan 1 & 2 + derivations, Aquila 1 & 2, Starcoder models, Refact,
|
||||||
|
Persimmon 8B, MPT, Bloom, Yi models, StableLM models, Deepseek models,
|
||||||
|
Qwen models, PLaMo-13B, Phi models, GPT-2, Orion 14B, InternLM2,
|
||||||
|
CodeShell, Gemma, Mamba, Grok-1, Xverse, Command-R models, SEA-LION,
|
||||||
|
GritLM-7B + GritLM-8x7B, OLMo, GPT-NeoX + Pythia, Snowflake-Arctic
|
||||||
|
MoE, Smaug, Poro 34B, Bitnet b1.58 models, Flan T5, Open Elm models,
|
||||||
|
ChatGLM3-6b + ChatGLM4-9b + GLMEdge-1.5b + GLMEdge-4b, SmolLM,
|
||||||
|
EXAONE-3.0-7.8B-Instruct, FalconMamba Models, Jais, Bielik-11B-v2.3,
|
||||||
|
RWKV-6, QRWKV-6, GigaChat-20B-A3B, Trillion-7B-preview, Ling models,
|
||||||
|
LFM2 models, Hunyuan models, BailingMoeV2 (Ring/Ling 2.0) models
|
||||||
|
|
||||||
|
Multimodal models:
|
||||||
|
|
||||||
|
LLaVA 1.5 models, BakLLaVA, Obsidian, ShareGPT4V, MobileVLM 1.7B/3B
|
||||||
|
models, Yi-VL, Mini CPM, Moondream, Bunny, GLM-EDGE, Qwen2-VL,
|
||||||
|
LFM2-VL
|
||||||
|
|
||||||
|
NOTE:
|
||||||
|
MODELS ARE NOT PROVIDED. You'll need to download them from the original
|
||||||
|
sites (or Hugging Face Hub).
|
||||||
|
|
||||||
|
Overall this is all raw and EXPERIMENTAL, no warranty, no support.
|
||||||
|
|
||||||
|
%package -n libllama-turboquant
|
||||||
|
Summary: Shared libraries for llama-cpp-turboquant
|
||||||
|
Group: System/Libraries
|
||||||
|
|
||||||
|
%description -n libllama-turboquant
|
||||||
|
%summary.
|
||||||
|
|
||||||
|
%package -n libllama-turboquant-devel
|
||||||
|
Summary: Development files for llama.cpp
|
||||||
|
Group: Development/C
|
||||||
|
Requires: libllama = %EVR
|
||||||
|
Conflicts: libwhisper-cpp-devel
|
||||||
|
|
||||||
|
%description -n libllama-turboquant-devel
|
||||||
|
%summary.
|
||||||
|
|
||||||
|
%package cpu
|
||||||
|
Summary: %name tools including backend for CPU
|
||||||
|
Group: Sciences/Computer science
|
||||||
|
Requires: libllama-turboquant = %EVR
|
||||||
|
Conflicts: %name-convert < %EVR
|
||||||
|
AutoReqProv: nopython3
|
||||||
|
%add_findreq_skiplist %_datadir/%name/examples/*
|
||||||
|
|
||||||
|
%description cpu
|
||||||
|
%summary.
|
||||||
|
|
||||||
|
%package cuda
|
||||||
|
Summary: %name backend for NVIDIA GPU
|
||||||
|
Group: Sciences/Computer science
|
||||||
|
Requires: libnvidia-ptxjitcompiler
|
||||||
|
Requires: %name-cpu = %EVR
|
||||||
|
|
||||||
|
%description cuda
|
||||||
|
%summary.
|
||||||
|
|
||||||
|
%package vulkan
|
||||||
|
Summary: %name backend for GPU
|
||||||
|
Group: Sciences/Computer science
|
||||||
|
Requires: %name-cpu = %EVR
|
||||||
|
|
||||||
|
%description vulkan
|
||||||
|
%summary.
|
||||||
|
|
||||||
|
%prep
|
||||||
|
%setup
|
||||||
|
%autopatch -p1
|
||||||
|
commit=$(awk '$2=="b%version"{print$1}' .gear/tags/list)
|
||||||
|
cat <<-EOF >> cmake/build-info.cmake
|
||||||
|
set(BUILD_NUMBER %version)
|
||||||
|
set(GGML_BUILD_NUMBER %version)
|
||||||
|
set(BUILD_COMMIT "${commit::8} [%release]")
|
||||||
|
EOF
|
||||||
|
sed -i '/POSITION_INDEPENDENT_CODE/s/PROPERTIES/& SOVERSION 0.0.%version/' src/CMakeLists.txt
|
||||||
|
sed -i 's/POSITION_INDEPENDENT_CODE/SOVERSION 0.0.%version &/' tools/mtmd/CMakeLists.txt
|
||||||
|
# We do not have Internet access (issues/13371).
|
||||||
|
perl -00 -ni -e 'print unless /_URL/' tests/test-arg-parser.cpp
|
||||||
|
# This test requires GPU.
|
||||||
|
sed /test-thread-safety/d -i tests/CMakeLists.txt
|
||||||
|
|
||||||
|
%build
|
||||||
|
%define optflags_debug -g1
|
||||||
|
# Unless -DCMAKE_SKIP_BUILD_RPATH=yes CMake fails to strip build time RPATH
|
||||||
|
# from (installed) binaries.
|
||||||
|
export NVCC_PREPEND_FLAGS=-ccbin=g++-12
|
||||||
|
%cmake \
|
||||||
|
-DCMAKE_SKIP_BUILD_RPATH=yes \
|
||||||
|
-DLLAMA_BUILD_TESTS=ON \
|
||||||
|
-DGGML_BACKEND_DL=ON \
|
||||||
|
-DGGML_BACKEND_DIR=%_libexecdir/llama-tq \
|
||||||
|
-DGGML_CPU=ON \
|
||||||
|
-DGGML_RPC=ON \
|
||||||
|
%ifarch x86_64
|
||||||
|
-DGGML_CPU_ALL_VARIANTS=ON \
|
||||||
|
%endif
|
||||||
|
%if_with cuda
|
||||||
|
-DGGML_CUDA=ON \
|
||||||
|
-DCMAKE_CUDA_ARCHITECTURES='52-virtual;80-virtual' \
|
||||||
|
%endif
|
||||||
|
%if_with vulkan
|
||||||
|
-DGGML_VULKAN=ON \
|
||||||
|
%endif
|
||||||
|
%nil
|
||||||
|
grep -E 'LLAMA|GGML' %_cmake__builddir/CMakeCache.txt | sort | tee build-options.txt
|
||||||
|
%cmake_build
|
||||||
|
find -name '*.py' | xargs sed -i '1s|#!/usr/bin/env python3|#!%__python3|'
|
||||||
|
LD_LIBRARY_PATH=%_cmake__builddir/bin %_cmake__builddir/bin/llama-tq-server --completion-bash > llama-tq.bash
|
||||||
|
LD_LIBRARY_PATH=%_cmake__builddir/bin .gear/gen-manpage %_cmake__builddir/bin/llama-tq-server > llama-tq-server.1
|
||||||
|
|
||||||
|
%install
|
||||||
|
%cmake_install
|
||||||
|
# Python requirements files.
|
||||||
|
install -Dpm644 requirements.txt -t %buildroot%_datadir/%name
|
||||||
|
cp -a requirements -t %buildroot%_datadir/%name
|
||||||
|
# Additional data.
|
||||||
|
cp -rp grammars -t %buildroot%_datadir/%name
|
||||||
|
# Not all examples.
|
||||||
|
install -Dp examples/*.sh -t %buildroot%_datadir/%name/examples
|
||||||
|
install -Dp examples/*.py -t %buildroot%_datadir/%name/examples
|
||||||
|
# We need to run the tests, not install them.
|
||||||
|
rm %buildroot%_bindir/test-*
|
||||||
|
rm %buildroot%_bindir/export-graph-ops
|
||||||
|
# Completions.
|
||||||
|
install -Dpm644 llama.bash %buildroot%_datadir/bash-completion/completions/llama-server
|
||||||
|
printf '%%s\n' llama-tq-cli llama-tq-simple llama-tq-run llama-tq-mtmd-cli |
|
||||||
|
xargs -ti ln -s llama-tq-server %buildroot%_datadir/bash-completion/completions/{}
|
||||||
|
mv %buildroot%_bindir/rpc-server %buildroot%_bindir/llama-tq-rpc-server
|
||||||
|
install -Dpm644 llama-tq-server.1 -t %buildroot%_man1dir
|
||||||
|
|
||||||
|
%check
|
||||||
|
( ! cuobjdump --list-elf %buildroot%_libexecdir/llama-tq/libggml-cuda.so | grep -F -v -e .cubin )
|
||||||
|
( ! cuobjdump --list-ptx %buildroot%_libexecdir/llama-tq/libggml-cuda.so | grep -F -v -e .sm_80.ptx -e .sm_52.ptx )
|
||||||
|
# Local path are more useful for debugging becasue they are not stripped by default.
|
||||||
|
export LD_LIBRARY_PATH=$PWD/%_cmake__builddir/bin PATH+=:$PWD/%_cmake__builddir/bin
|
||||||
|
llama-tq-server --version
|
||||||
|
llama-tq-server --version |& grep -Ex 'version: %version \(\S+ \[%release\]\)'
|
||||||
|
# test-eval-callback wants network.
|
||||||
|
%ctest -E 'test-download-model|test-eval-callback|test-state-restore-fragmented|test-llama-archs'
|
||||||
|
# It broke so it can suck my dick
|
||||||
|
#llama-completion -m /usr/share/tinyllamas/stories260K.gguf -p "Hello" -s 42 -n 500 2>/dev/null
|
||||||
|
#llama-completion -m /usr/share/tinyllamas/stories260K.gguf -p "Once upon a time" -s 55 -n 33 2>/dev/null |
|
||||||
|
# grep 'Once upon a time, there was a boy named Tom. Tom had a big box of colors.'
|
||||||
|
# We do not provide convert tools.
|
||||||
|
mv %buildroot%_bindir/convert*.py -t %buildroot%_datadir/%name/examples
|
||||||
|
|
||||||
|
%files
|
||||||
|
|
||||||
|
%files -n libllama-turboquant
|
||||||
|
%_libdir/libllama-tq.so.0.0.%version
|
||||||
|
%_libdir/libggml-tq.so.0
|
||||||
|
%_libdir/libggml-tq.so.0.*
|
||||||
|
%_libdir/libggml-tq-base.so.0
|
||||||
|
%_libdir/libggml-tq-base.so.0.*
|
||||||
|
%_libdir/libmtmd-tq.so.0.0.%version
|
||||||
|
|
||||||
|
%files -n libllama-turboquant-devel
|
||||||
|
%_libdir/libllama-tq.so
|
||||||
|
%_libdir/libggml-tq.so
|
||||||
|
%_libdir/libggml-tq-base.so
|
||||||
|
%_libdir/libmtmd-tq.so
|
||||||
|
%_includedir/llama-tq*.h
|
||||||
|
%_includedir/gguf-tq.h
|
||||||
|
%_includedir/ggml-tq*.h
|
||||||
|
%_includedir/mtmd-tq*.h
|
||||||
|
%_cmakedir/ggml-tq
|
||||||
|
%_cmakedir/llama-tq
|
||||||
|
%_pkgconfigdir/llama-tq.pc
|
||||||
|
|
||||||
|
%files cpu
|
||||||
|
%define _customdocdir %_docdir/%name
|
||||||
|
%doc LICENSE README.md docs build-options.txt
|
||||||
|
%_bindir/llama-tq-*
|
||||||
|
%dir %_datadir/%name
|
||||||
|
%dir %_datadir/%name/examples
|
||||||
|
%_datadir/%name/examples/*.sh
|
||||||
|
%_datadir/%name/examples/*.py
|
||||||
|
%_datadir/%name/requirements*
|
||||||
|
%_datadir/%name/grammars
|
||||||
|
%dir %_libexecdir/llama-tq
|
||||||
|
%_libexecdir/llama-tq/libggml-cpu*.so
|
||||||
|
%_libexecdir/llama-tq/libggml-rpc.so
|
||||||
|
%_datadir/bash-completion/completions/llama-tq-*
|
||||||
|
%_man1dir/llama-tq-server.1*
|
||||||
|
|
||||||
|
%if_with cuda
|
||||||
|
%files cuda
|
||||||
|
%dir %_libexecdir/llama
|
||||||
|
%_libexecdir/llama-tq/libggml-cuda.so
|
||||||
|
%endif
|
||||||
|
|
||||||
|
%if_with vulkan
|
||||||
|
%files vulkan
|
||||||
|
%dir %_libexecdir/llama-tq
|
||||||
|
%_libexecdir/llama-tq/libggml-vulkan.so
|
||||||
|
%endif
|
||||||
|
|
||||||
|
%changelog
|
||||||
|
* Tue Apr 07 2026 Arseniy Romenskiy <romenskiy@altlinux.org> 1:8659-alt1
|
||||||
|
- Update to b8659 (2026-04-07).
|
||||||
|
|
||||||
|
* Mon Apr 06 2026 Vitaly Chikunov <vt@altlinux.org> 1:8681-alt1
|
||||||
|
- Update to b8681 (2026-04-06).
|
||||||
|
|
||||||
|
* Sun Mar 22 2026 Vitaly Chikunov <vt@altlinux.org> 1:8470-alt1
|
||||||
|
- Update to b8470 (2026-03-22).
|
||||||
|
|
||||||
|
* Tue Mar 03 2026 Vitaly Chikunov <vt@altlinux.org> 1:8192-alt1
|
||||||
|
- Update to b8192 (2026-03-03).
|
||||||
|
|
||||||
|
* Fri Feb 13 2026 Vitaly Chikunov <vt@altlinux.org> 1:8018-alt1
|
||||||
|
- Update to b8018 (2026-02-12).
|
||||||
|
|
||||||
|
* Sat Jan 24 2026 Vitaly Chikunov <vt@altlinux.org> 1:7819-alt1
|
||||||
|
- Update to b7819 (2026-01-23).
|
||||||
|
- Responses API support (partial).
|
||||||
|
|
||||||
|
* Sun Dec 14 2025 Vitaly Chikunov <vt@altlinux.org> 1:7388-alt1
|
||||||
|
- Update to b7388 (2025-12-13).
|
||||||
|
- llama-cli: New CLI experience (with the old moved to llama-completion).
|
||||||
|
- llama-server: Live model switching.
|
||||||
|
- Messages API support.
|
||||||
|
|
||||||
|
* Fri Nov 21 2025 Vitaly Chikunov <vt@altlinux.org> 1:7127-alt1
|
||||||
|
- Update to b7127 (2025-11-21).
|
||||||
|
- spec: Remove llama.cpp-convert package.
|
||||||
|
- model: detect GigaChat3-10-A1.8B as deepseek lite.
|
||||||
|
|
||||||
|
* Tue Oct 28 2025 Vitaly Chikunov <vt@altlinux.org> 1:6869-alt1
|
||||||
|
- Update to b6869 (2025-10-28).
|
||||||
|
|
||||||
|
* Sat Sep 06 2025 Vitaly Chikunov <vt@altlinux.org> 1:6397-alt1
|
||||||
|
- Update to b6397 (2025-09-06).
|
||||||
|
- Python-based model conversion scripts are sub-packaged. Note that they are
|
||||||
|
not supported and are provided as-is.
|
||||||
|
|
||||||
|
* Sat Aug 09 2025 Vitaly Chikunov <vt@altlinux.org> 1:6121-alt1
|
||||||
|
- Update to b6121 (2025-08-08).
|
||||||
|
|
||||||
|
* Wed Jun 25 2025 Vitaly Chikunov <vt@altlinux.org> 1:5753-alt1
|
||||||
|
- Update to b5753 (2025-06-24).
|
||||||
|
- Install an experimental rpc backend and server. The rpc code is a
|
||||||
|
proof-of-concept, fragile, and insecure.
|
||||||
|
|
||||||
|
* Sat May 10 2025 Vitaly Chikunov <vt@altlinux.org> 1:5332-alt1
|
||||||
|
- Update to b5332 (2025-05-09), with vision support in llama-server.
|
||||||
|
- Enable Vulkan backend (for GPU) in llama.cpp-vulkan package.
|
||||||
|
|
||||||
|
* Mon Mar 10 2025 Vitaly Chikunov <vt@altlinux.org> 1:4855-alt1
|
||||||
|
- Update to b4855 (2025-03-07).
|
||||||
|
- Enable CUDA backend (for NVIDIA GPU) in llama.cpp-cuda package.
|
||||||
|
- Disable BLAS backend (issues/12282).
|
||||||
|
- Install bash-completions.
|
||||||
|
|
||||||
|
* Tue Jul 23 2024 Vitaly Chikunov <vt@altlinux.org> 1:3441-alt1
|
||||||
|
- Update to b3441 (2024-07-23).
|
||||||
|
- spec: Package libllama and libllama-devel (ALT#50962).
|
||||||
|
- spec: Use upstream install procedure; as a consequence, some binary names are
|
||||||
|
changed.
|
||||||
|
|
||||||
|
* Mon Jun 03 2024 Vitaly Chikunov <vt@altlinux.org> 1:3072-alt1.20240603
|
||||||
|
- Update to b3072 (2024-06-03).
|
||||||
|
- The version scheme now matches the upstream build number more closely,
|
||||||
|
instead of using the commit date.
|
||||||
|
- Build with libcurl and OpenBLAS support.
|
||||||
|
|
||||||
|
* Tue May 28 2024 Vitaly Chikunov <vt@altlinux.org> 20240527-alt1
|
||||||
|
- Update to b3012 (2024-05-27).
|
||||||
|
|
||||||
|
* Mon Feb 26 2024 Vitaly Chikunov <vt@altlinux.org> 20240225-alt1
|
||||||
|
- Update to b2259 (2024-02-25).
|
||||||
|
|
||||||
|
* Fri Oct 20 2023 Vitaly Chikunov <vt@altlinux.org> 20231019-alt1
|
||||||
|
- Update to b1400 (2023-10-19).
|
||||||
|
- Install experimental converters (convert- prefixed tools).
|
||||||
|
|
||||||
|
* Sun Jul 30 2023 Vitaly Chikunov <vt@altlinux.org> 20230728-alt1
|
||||||
|
- Update to master-8a88e58 (2023-07-28).
|
||||||
|
|
||||||
|
* Sun May 14 2023 Vitaly Chikunov <vt@altlinux.org> 20230513-alt1
|
||||||
|
- Build master-bda4d7c (2023-05-13).
|
||||||
|
|
||||||
|
* Wed Apr 19 2023 Vitaly Chikunov <vt@altlinux.org> 20230419-alt1
|
||||||
|
- Build master-6667401 (2023-04-19).
|
||||||
14
.gear/llama.service
Normal file
14
.gear/llama.service
Normal file
|
|
@ -0,0 +1,14 @@
|
||||||
|
[Unit]
|
||||||
|
Description=Llama.cpp server, CPU only (no GPU support in this build).
|
||||||
|
After=syslog.target network.target local-fs.target remote-fs.target nss-lookup.target
|
||||||
|
|
||||||
|
[Service]
|
||||||
|
Type=simple
|
||||||
|
DynamicUser=true
|
||||||
|
EnvironmentFile=-/etc/sysconfig/llama
|
||||||
|
ExecStart=llama-server $LLAMA_ARGS
|
||||||
|
ExecReload=/bin/kill -HUP $MAINPID
|
||||||
|
Restart=never
|
||||||
|
|
||||||
|
[Install]
|
||||||
|
WantedBy=default.target
|
||||||
5
.gear/npm-build
Executable file
5
.gear/npm-build
Executable file
|
|
@ -0,0 +1,5 @@
|
||||||
|
workdir=tools/server/webui
|
||||||
|
target=tools/server/public
|
||||||
|
aikido-npm ci --ignore-scripts
|
||||||
|
aikido-npm audit --audit-level=critical fix
|
||||||
|
npm run build
|
||||||
5
.gear/rules
Normal file
5
.gear/rules
Normal file
|
|
@ -0,0 +1,5 @@
|
||||||
|
# The diff is only to detect commit hash of release tag `gear-update-tag -ac`
|
||||||
|
# for version output.
|
||||||
|
tar: b@version@:.
|
||||||
|
diff: b@version@:. . name=@name@-@version@.patch
|
||||||
|
spec: .gear/llama-cpp-turboquant.spec
|
||||||
1
.gear/tags/list
Normal file
1
.gear/tags/list
Normal file
|
|
@ -0,0 +1 @@
|
||||||
|
e43970099269b5b6da36b8977ad47697602e4e54 b8659
|
||||||
41
.gear/upstream-signing-key.asc
Normal file
41
.gear/upstream-signing-key.asc
Normal file
|
|
@ -0,0 +1,41 @@
|
||||||
|
-----BEGIN PGP PUBLIC KEY BLOCK-----
|
||||||
|
|
||||||
|
mQENBFmUaEEBCACzXTDt6ZnyaVtueZASBzgnAmK13q9Urgch+sKYeIhdymjuMQta
|
||||||
|
x15OklctmrZtqre5kwPUosG3/B2/ikuPYElcHgGPL4uL5Em6S5C/oozfkYzhwRrT
|
||||||
|
SQzvYjsE4I34To4UdE9KA97wrQjGoz2Bx72WDLyWwctD3DKQtYeHXswXXtXwKfjQ
|
||||||
|
7Fy4+Bf5IPh76dA8NJ6UtjjLIDlKqdxLW4atHe6xWFaJ+XdLUtsAroZcXBeWDCPa
|
||||||
|
buXCDscJcLJRKZVc62gOZXXtPfoHqvUPp3nuLA4YjH9bphbrMWMf810Wxz9JTd3v
|
||||||
|
yWgGqNY0zbBqeZoGv+TuExlRHT8ASGFS9SVDABEBAAG0NUdpdEh1YiAod2ViLWZs
|
||||||
|
b3cgY29tbWl0IHNpZ25pbmcpIDxub3JlcGx5QGdpdGh1Yi5jb20+iQEoBBMBCAAc
|
||||||
|
BQJZlGhBCRBK7hj4Ov3rIwIbAwUJDBJ3/wIZAQAA0O4IAJd0k8M+urETyMvTqNTj
|
||||||
|
/U6nbqyOdKE4V93uUj5G7sNTfno7wod/Qjj6Zv5KodvA93HmEdQqsmVq5YJ5KGiw
|
||||||
|
cmGCpd/GqJRPaYSY0hSUSBqYHiHLusCJkPBpQTBhcEMtfVCB2J6fVeoX2DV0K1xf
|
||||||
|
CGblrSVB0viAxUMnmL5C55RuvbYZsTu8szXhkvIR96CtWbJ8QGaEf1/KSpWz8ept
|
||||||
|
Y/omf3UPfvdOjnsxc8jVEqPNaR9xC6Q6t53rBa/XgMY6IYyesnyYnc5O6JuexUFa
|
||||||
|
VjykRFtAiYfDaMARpXOmgMm0lhoBRKb/uMUaN3CSYTmE4pZweJcUi7eWgmoQljX2
|
||||||
|
ut6ZAg0EZabFdgEQALI37i+IVAzpBCgqvQDZbSsZ0yhtMnA5myjZA+l7BvIGy4ve
|
||||||
|
s1bk6YetbBcCE8o2pQjI7N2rwyhLGhNO6ouSyhqGLEQv9fafKE4HFH0aRjP+gj1H
|
||||||
|
edhwtFoVChImhV863rWimQtTNtYB6GluBPwQqWfwmwQ2rT7ScOVZCLSHZD2gaaqW
|
||||||
|
BXOyTCZVnwt7K/gyDuE3qzDJnuahl+SSkPn5TtnZdW6sLORJJ+DjNvaUxEsmizZ4
|
||||||
|
IBzvj0QKxfS3s4F+0X5iqCMheLFeybZGtSq9Tjs6Q61l4CG8Bh6dsLemv0WFrk3G
|
||||||
|
gFQRr7XUwr1bo5xGHC/FUJSsxRHoVNJnIL/9WldNO2tGU6qlTnAYxs/fOmf2B6o5
|
||||||
|
cKXysXv7WAA8b+j5AVBMGxUSu7CLglaiCJC5DI7AAiUV7/t29rFZkam//Jbb4veC
|
||||||
|
4vvFocoVUaxrKGWK1BDldr4/WJKApJcPJF4Jtai1+oB6ak/JIjbkseHdJxcjo2B0
|
||||||
|
dKtIFoWiPAB+DFs9MRDpp0iwocJCh+ucus1rdQ54YMaI44rRphXeOIQMYCi5q2Q1
|
||||||
|
/arzkSiyPV/2VoKoAfdgskPt1xKd7WIKErmpFMHIy8jJ5IPQ1s2dUwU4alfJLJa0
|
||||||
|
pvaV2m7wBYFAmwmz0WZgFxYAYEDamn4jFoKfqsEgcixRUVE3w5VkqwSwGRbLABEB
|
||||||
|
AAG0G0dpdEh1YiA8bm9yZXBseUBnaXRodWIuY29tPokCTgQTAQoAOBYhBJaEeaGv
|
||||||
|
+SfjfRpWa7VpDu67lSGUBQJlpsV2AhsDBQsJCAcCBhUKCQgLAgQWAgMBAh4BAheA
|
||||||
|
AAoJELVpDu67lSGUgy4QAKW9XAL416iKrQB2LElmxqAoenHVCswlau0xGLh5dVNN
|
||||||
|
p5f4/W6eEL8CZI7hfF3e5Gh6Me99aHgXSCK1QnxcqCJ6Oea4ZyrsNu3k6g7Um5ca
|
||||||
|
VbYFD4yIahhXDYHSw6FYM2sgFY479YvgvKRwacC2tFfChLRbHgwLJ3O1dBjmVycJ
|
||||||
|
Zpbyu+7taZ26g6KQfgcj3uuo3nz3p1ziIEpLHwtl/7joNEIIP/lJ8AKmUHPiGznN
|
||||||
|
6fxMvzN37PGMWtdvOi1rSNIMQYr1YY7jPnlLbFJwLrO/q/cGPU5HwGzlqh0a2ZqY
|
||||||
|
dnuwT3DREmgJ83H71xH+sTzZKs5oGlVTu6st7iWDvNpo2GoN01XzKa5caYglqsOC
|
||||||
|
uZ6IHlsdL50sXMtSROCi3hEWU9r1sWIm4k3pNz20y7lElD2X/MqbEMcgpawCV7lH
|
||||||
|
rm7MSrTgu6BNAF0SisbF9AKwXaBr2dwpMMyIBOFZO9mk4/c0n9q2FlGY4GkbgH2J
|
||||||
|
HqulFTwX/4yiQbh8gzCe+06FZAWITN1OQntTkkCQ+1MCZPf+bOfC08RTsOsVZIYB
|
||||||
|
2qAgw6XE0IF4a+PAtHSoYftwH2ocMY2gMuSNpQWm7m0+/j+K+RBoeUcnGNPQgszq
|
||||||
|
N60IDMqkqHjyubrm2aslfopWmPSvaQoyxwV/uztdo+UI0IV2z9gD7Sm49vMkpYp8
|
||||||
|
=QeVw
|
||||||
|
-----END PGP PUBLIC KEY BLOCK-----
|
||||||
Loading…
Add table
Add a link
Reference in a new issue