Skip to content

GitLab

Unverified Commit 652c273f authored by Daniel Hiltgen's avatar Daniel Hiltgen Committed by GitHub
Browse files

Merge pull request #5049 from dhiltgen/cuda_v12 

Cuda v12
parents 88e77050 f9e31da9
Hide whitespace changes
Inline Side-by-side
Showing with 101 additions and 27 deletions
scripts/build_windows.ps1
View file @ 652c273f
......@@ -7,6 +7,7 @@
$ErrorActionPreference = "Stop"
function checkEnv() {
$script:ARCH = $Env:PROCESSOR_ARCHITECTURE.ToLower()
$script:TARGET_ARCH=$Env:PROCESSOR_ARCHITECTURE.ToLower()
Write-host "Building for ${script:TARGET_ARCH}"
write-host "Locating required tools and paths"
......@@ -15,26 +16,23 @@ function checkEnv() {
$MSVC_INSTALL=(Get-CimInstance MSFT_VSInstance -Namespace root/cimv2/vs)[0].InstallLocation
$env:VCToolsRedistDir=(get-item "${MSVC_INSTALL}\VC\Redist\MSVC\*")[0]
}
# Try to find the CUDA dir
if ($null -eq $env:NVIDIA_DIR) {
# Locate CUDA versions
# Note: this assumes every version found will be built
$cudaList=(get-item "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v*\bin\" -ea 'silentlycontinue')
if ($cudaList.length -eq 0) {
$d=(get-command -ea 'silentlycontinue' nvcc).path
if ($d -ne $null) {
$script:NVIDIA_DIR=($d| split-path -parent)
} else {
$cudaList=(get-item "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v*\bin\" -ea 'silentlycontinue')
if ($cudaList.length > 0) {
$script:NVIDIA_DIR=$cudaList[0]
}
if ($null -ne $d) {
$script:CUDA_DIRS=@($d| split-path -parent)
}
} else {
$script:NVIDIA_DIR=$env:NVIDIA_DIR
$script:CUDA_DIRS=$cudaList
}
$script:INNO_SETUP_DIR=(get-item "C:\Program Files*\Inno Setup*\")[0]
$script:DEPS_DIR="${script:SRC_DIR}\dist\windows-${script:TARGET_ARCH}"
$env:CGO_ENABLED="1"
echo "Checking version"
Write-Output "Checking version"
if (!$env:VERSION) {
$data=(git describe --tags --first-parent --abbrev=7 --long --dirty --always)
$pattern="v(.+)"
......@@ -71,7 +69,48 @@ function checkEnv() {
function buildOllama() {
write-host "Building ollama CLI"
if ($null -eq ${env:OLLAMA_SKIP_GENERATE}) {
& go generate ./...
Remove-Item -ea 0 -recurse -force -path "${script:SRC_DIR}\dist\windows-${script:ARCH}"
# TODO - consider trying to parallelize this with Start-ThreadJob, but env vars can't be used to toggle
# which targets to build
# Start by skipping CUDA to build everything else
pwsh -Command { $env:OLLAMA_SKIP_CUDA_GENERATE="1"; & go generate ./... }
if ($LASTEXITCODE -ne 0) { exit($LASTEXITCODE)}
# Then skip everyhting else and build all the CUDA variants
foreach ($env:CUDA_LIB_DIR in $script:CUDA_DIRS) {
write-host "Building CUDA ${env:CUDA_LIB_DIR}"
if ($env:CUDA_LIB_DIR.Contains("v12")) {
pwsh -Command {
$env:OLLAMA_SKIP_CUDA_GENERATE=""
$env:OLLAMA_SKIP_STATIC_GENERATE="1"
$env:OLLAMA_SKIP_CPU_GENERATE="1"
$env:OLLAMA_SKIP_ONEAPI_GENERATE="1"
$env:OLLAMA_SKIP_ROCM_GENERATE="1"
$env:CMAKE_CUDA_ARCHITECTURES="60;61;62;70;72;75;80;86;87;89;90;90a"
$env:OLLAMA_CUSTOM_CUDA_DEFS="-DGGML_CUDA_USE_GRAPHS=on"
$env:CUDA_PATH=split-path -path $env:CUDA_LIB_DIR -parent
$env:PATH="$envs:CUDA_LIB_DIR;$env:PATH"
& go generate ./...
}
} else {
pwsh -Command {
$env:OLLAMA_SKIP_CUDA_GENERATE=""
$env:OLLAMA_SKIP_STATIC_GENERATE="1"
$env:OLLAMA_SKIP_CPU_GENERATE="1"
$env:OLLAMA_SKIP_ONEAPI_GENERATE="1"
$env:OLLAMA_SKIP_ROCM_GENERATE="1"
$env:CMAKE_CUDA_ARCHITECTURES="50;52;53;60;61;62;70;72;75;80;86"
$env:OLLAMA_CUSTOM_CUDA_DEFS=""
$env:CUDA_PATH=split-path -path $env:CUDA_LIB_DIR -parent
$env:PATH="$envs:CUDA_LIB_DIR;$env:PATH"
& go generate ./...
}
}
if ($LASTEXITCODE -ne 0) { exit($LASTEXITCODE)}
}
if ($LASTEXITCODE -ne 0) { exit($LASTEXITCODE)}
} else {
write-host "Skipping generate step with OLLAMA_SKIP_GENERATE set"
......@@ -83,8 +122,8 @@ function buildOllama() {
/csp "Google Cloud KMS Provider" /kc ${env:KEY_CONTAINER} ollama.exe
if ($LASTEXITCODE -ne 0) { exit($LASTEXITCODE)}
}
New-Item -ItemType Directory -Path .\dist\windows-${script:TARGET_ARCH}\ -Force
cp .\ollama.exe .\dist\windows-${script:TARGET_ARCH}\
New-Item -ItemType Directory -Path .\dist\windows-${script:TARGET_ARCH}\bin\ -Force
cp .\ollama.exe .\dist\windows-${script:TARGET_ARCH}\bin\
}
function buildApp() {
......@@ -103,22 +142,22 @@ function buildApp() {
function gatherDependencies() {
write-host "Gathering runtime dependencies"
cd "${script:SRC_DIR}"
md "${script:DEPS_DIR}\ollama_runners" -ea 0 > $null
md "${script:DEPS_DIR}\lib\ollama" -ea 0 > $null
# TODO - this varies based on host build system and MSVC version - drive from dumpbin output
# currently works for Win11 + MSVC 2019 + Cuda V11
cp "${env:VCToolsRedistDir}\x64\Microsoft.VC*.CRT\msvcp140*.dll" "${script:DEPS_DIR}\ollama_runners\"
cp "${env:VCToolsRedistDir}\x64\Microsoft.VC*.CRT\vcruntime140.dll" "${script:DEPS_DIR}\ollama_runners\"
cp "${env:VCToolsRedistDir}\x64\Microsoft.VC*.CRT\vcruntime140_1.dll" "${script:DEPS_DIR}\ollama_runners\"
cp "${env:VCToolsRedistDir}\x64\Microsoft.VC*.CRT\msvcp140*.dll" "${script:DEPS_DIR}\lib\ollama\"
cp "${env:VCToolsRedistDir}\x64\Microsoft.VC*.CRT\vcruntime140.dll" "${script:DEPS_DIR}\lib\ollama\"
cp "${env:VCToolsRedistDir}\x64\Microsoft.VC*.CRT\vcruntime140_1.dll" "${script:DEPS_DIR}\lib\ollama\"
foreach ($part in $("runtime", "stdio", "filesystem", "math", "convert", "heap", "string", "time", "locale", "environment")) {
cp "$env:VCToolsRedistDir\..\..\..\Tools\Llvm\x64\bin\api-ms-win-crt-${part}*.dll" "${script:DEPS_DIR}\ollama_runners\"
cp "$env:VCToolsRedistDir\..\..\..\Tools\Llvm\x64\bin\api-ms-win-crt-${part}*.dll" "${script:DEPS_DIR}\lib\ollama\"
}
cp "${script:SRC_DIR}\app\ollama_welcome.ps1" "${script:SRC_DIR}\dist\"
if ("${env:KEY_CONTAINER}") {
write-host "about to sign"
foreach ($file in (get-childitem "${script:DEPS_DIR}\cuda\cu*.dll") + @("${script:SRC_DIR}\dist\ollama_welcome.ps1")){
foreach ($file in (get-childitem "${script:DEPS_DIR}\lib\ollama\cu*.dll") + @("${script:SRC_DIR}\dist\ollama_welcome.ps1")){
write-host "signing $file"
& "${script:SignTool}" sign /v /fd sha256 /t http://timestamp.digicert.com /f "${script:OLLAMA_CERT}" `
/csp "Google Cloud KMS Provider" /kc ${env:KEY_CONTAINER} $file
......
scripts/install.sh
View file @ 652c273f
......@@ -63,16 +63,36 @@ if [ -n "$NEEDS" ]; then
exit 1
fi
status "Downloading ollama..."
curl --fail --show-error --location --progress-bar -o $TEMP_DIR/ollama "https://ollama.com/download/ollama-linux-${ARCH}${VER_PARAM}"
for BINDIR in /usr/local/bin /usr/bin /bin; do
echo $PATH | grep -q $BINDIR && break || continue
done
OLLAMA_INSTALL_DIR=$(dirname ${BINDIR})
status "Installing ollama to $BINDIR..."
status "Installing ollama to $OLLAMA_INSTALL_DIR"
$SUDO install -o0 -g0 -m755 -d $BINDIR
$SUDO install -o0 -g0 -m755 $TEMP_DIR/ollama $BINDIR/ollama
$SUDO install -o0 -g0 -m755 -d "$OLLAMA_INSTALL_DIR"
if curl -I --silent --fail --location "https://ollama.com/download/ollama-linux-${ARCH}.tgz${VER_PARAM}" >/dev/null ; then
status "Downloading Linux ${ARCH} bundle"
curl --fail --show-error --location --progress-bar \
"https://ollama.com/download/ollama-linux-${ARCH}.tgz${VER_PARAM}" | \
$SUDO tar -xzf - -C "$OLLAMA_INSTALL_DIR"
BUNDLE=1
if [ "$OLLAMA_INSTALL_DIR/bin/ollama" != "$BINDIR/ollama" ] ; then
status "Making ollama accessible in the PATH in $BINDIR"
$SUDO ln -sf "$OLLAMA_INSTALL_DIR/ollama" "$BINDIR/ollama"
fi
else
status "Downloading Linux ${ARCH} CLI"
curl --fail --show-error --location --progress-bar -o "$TEMP_DIR/ollama"\
"https://ollama.com/download/ollama-linux-${ARCH}${VER_PARAM}"
$SUDO install -o0 -g0 -m755 $TEMP_DIR/ollama $OLLAMA_INSTALL_DIR/ollama
BUNDLE=0
if [ "$OLLAMA_INSTALL_DIR/ollama" != "$BINDIR/ollama" ] ; then
status "Making ollama accessible in the PATH in $BINDIR"
$SUDO ln -sf "$OLLAMA_INSTALL_DIR/ollama" "$BINDIR/ollama"
fi
fi
install_success() {
status 'The Ollama API is now available at 127.0.0.1:11434.'
......@@ -178,6 +198,11 @@ if ! check_gpu lspci nvidia && ! check_gpu lshw nvidia && ! check_gpu lspci amdg
fi
if check_gpu lspci amdgpu || check_gpu lshw amdgpu; then
if [ $BUNDLE -ne 0 ]; then
install_success
status "AMD GPU ready."
exit 0
fi
# Look for pre-existing ROCm v6 before downloading the dependencies
for search in "${HIP_PATH:-''}" "${ROCM_PATH:-''}" "/opt/rocm" "/usr/lib64"; do
if [ -n "${search}" ] && [ -e "${search}/libhipblas.so.2" -o -e "${search}/lib/libhipblas.so.2" ]; then
......
scripts/rh_linux_deps.sh
View file @ 652c273f
......@@ -3,6 +3,7 @@
# Script for common Dockerfile dependency installation in redhat linux based images
set -ex
set -o pipefail
MACHINE=$(uname -m)
if grep -i "centos" /etc/system-release >/dev/null; then
......@@ -29,7 +30,7 @@ if grep -i "centos" /etc/system-release >/dev/null; then
dnf install -y rh-git227-git
ln -s /opt/rh/rh-git227/root/usr/bin/git /usr/local/bin/git
fi
dnf install -y devtoolset-10-gcc devtoolset-10-gcc-c++
dnf install -y devtoolset-10-gcc devtoolset-10-gcc-c++ pigz
elif grep -i "rocky" /etc/system-release >/dev/null; then
# Temporary workaround until rocky 8 AppStream ships GCC 10.4 (10.3 is incompatible with NVCC)
cat << EOF > /etc/yum.repos.d/Rocky-Vault.repo
......@@ -43,12 +44,21 @@ gpgkey=file:///etc/pki/rpm-gpg/RPM-GPG-KEY-rockyofficial
EOF
dnf install -y git \
gcc-toolset-10-gcc-10.2.1-8.2.el8 \
gcc-toolset-10-gcc-c++-10.2.1-8.2.el8
gcc-toolset-10-gcc-c++-10.2.1-8.2.el8 \
pigz
else
echo "ERROR Unexpected distro"
exit 1
fi
if [ "${MACHINE}" = "x86_64" ] ; then
curl -s -L https://github.com/ccache/ccache/releases/download/v4.10.2/ccache-4.10.2-linux-x86_64.tar.xz | tar -Jx -C /tmp --strip-components 1 && \
mv /tmp/ccache /usr/local/bin/
else
yum -y install epel-release
yum install -y ccache
fi
if [ -n "${CMAKE_VERSION}" ]; then
curl -s -L https://github.com/Kitware/CMake/releases/download/v${CMAKE_VERSION}/cmake-${CMAKE_VERSION}-linux-$(uname -m).tar.gz | tar -zx -C /usr --strip-components 1
fi
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment