CI : Enable CUDA and Vulkan ARM64 runners and fix CI/CD (#21122)

* CI: Enable CUDA and Vulkan ARM64 runners and fix CI/CD

Co-authored-by: Ts-sound <44093942+Ts-sound@users.noreply.github.com>

* Obtain source tag name from git tag

Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com>

---------

Co-authored-by: Ts-sound <44093942+Ts-sound@users.noreply.github.com>
Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com>
This commit is contained in:
Seungmin Kim 2026-03-31 03:24:37 +09:00 committed by GitHub
parent ead417f01c
commit 84ae8434d0
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
10 changed files with 436 additions and 205 deletions

View File

@ -36,7 +36,7 @@ RUN mkdir -p /app/full \
FROM ubuntu:$UBUNTU_VERSION AS base
RUN apt-get update \
&& apt-get install -y libgomp1 curl\
&& apt-get install -y libgomp1 curl \
&& apt autoremove -y \
&& apt clean -y \
&& rm -rf /tmp/* /var/tmp/* \

View File

@ -1,6 +1,6 @@
ARG UBUNTU_VERSION=24.04
# This needs to generally match the container host's environment.
ARG CUDA_VERSION=13.1.0
ARG CUDA_VERSION=13.1.1
# Target the CUDA build image
ARG BASE_CUDA_DEV_CONTAINER=nvidia/cuda:${CUDA_VERSION}-devel-ubuntu${UBUNTU_VERSION}
@ -12,7 +12,9 @@ FROM ${BASE_CUDA_DEV_CONTAINER} AS build
ARG CUDA_DOCKER_ARCH=default
RUN apt-get update && \
apt-get install -y build-essential cmake python3 python3-pip git libssl-dev libgomp1
apt-get install -y gcc-14 g++-14 build-essential cmake python3 python3-pip git libssl-dev libgomp1
ENV CC=gcc-14 CXX=g++-14 CUDAHOSTCXX=g++-14
WORKDIR /app
@ -39,7 +41,7 @@ RUN mkdir -p /app/full \
FROM ${BASE_CUDA_RUN_CONTAINER} AS base
RUN apt-get update \
&& apt-get install -y libgomp1 curl\
&& apt-get install -y libgomp1 curl \
&& apt autoremove -y \
&& apt clean -y \
&& rm -rf /tmp/* /var/tmp/* \

View File

@ -1,6 +1,6 @@
ARG UBUNTU_VERSION=22.04
ARG UBUNTU_VERSION=24.04
# This needs to generally match the container host's environment.
ARG CUDA_VERSION=12.4.0
ARG CUDA_VERSION=12.8.1
# Target the CUDA build image
ARG BASE_CUDA_DEV_CONTAINER=nvidia/cuda:${CUDA_VERSION}-devel-ubuntu${UBUNTU_VERSION}
@ -12,7 +12,9 @@ FROM ${BASE_CUDA_DEV_CONTAINER} AS build
ARG CUDA_DOCKER_ARCH=default
RUN apt-get update && \
apt-get install -y build-essential cmake python3 python3-pip git libssl-dev libgomp1
apt-get install -y gcc-14 g++-14 build-essential cmake python3 python3-pip git libssl-dev libgomp1
ENV CC=gcc-14 CXX=g++-14 CUDAHOSTCXX=g++-14
WORKDIR /app
@ -39,7 +41,7 @@ RUN mkdir -p /app/full \
FROM ${BASE_CUDA_RUN_CONTAINER} AS base
RUN apt-get update \
&& apt-get install -y libgomp1 curl\
&& apt-get install -y libgomp1 curl \
&& apt autoremove -y \
&& apt clean -y \
&& rm -rf /tmp/* /var/tmp/* \
@ -60,7 +62,8 @@ RUN apt-get update \
git \
python3 \
python3-pip \
&& pip install --upgrade pip setuptools wheel \
python3-wheel \
&& pip install --break-system-packages --upgrade setuptools \
&& pip install --break-system-packages -r requirements.txt \
&& apt autoremove -y \
&& apt clean -y \

View File

@ -51,7 +51,7 @@ RUN mkdir /tmp/neo/ && cd /tmp/neo/ \
&& dpkg --install *.deb
RUN apt-get update \
&& apt-get install -y libgomp1 curl\
&& apt-get install -y libgomp1 curl \
&& apt autoremove -y \
&& apt clean -y \
&& rm -rf /tmp/* /var/tmp/* \

View File

@ -46,7 +46,7 @@ RUN mkdir -p /app/full \
FROM ${BASE_MUSA_RUN_CONTAINER} AS base
RUN apt-get update \
&& apt-get install -y libgomp1 curl\
&& apt-get install -y libgomp1 curl \
&& apt autoremove -y \
&& apt clean -y \
&& rm -rf /tmp/* /var/tmp/* \

View File

@ -78,7 +78,7 @@ ARG http_proxy
ARG https_proxy
RUN apt-get update \
&& apt-get install -y libgomp1 libtbb12 curl\
&& apt-get install -y libgomp1 libtbb12 curl \
&& apt autoremove -y \
&& apt clean -y \
&& rm -rf /tmp/* /var/tmp/* \

View File

@ -58,7 +58,7 @@ RUN mkdir -p /app/full \
FROM ${BASE_ROCM_DEV_CONTAINER} AS base
RUN apt-get update \
&& apt-get install -y libgomp1 curl\
&& apt-get install -y libgomp1 curl \
&& apt autoremove -y \
&& apt clean -y \
&& rm -rf /tmp/* /var/tmp/* \
@ -79,7 +79,7 @@ RUN apt-get update \
git \
python3-pip \
python3 \
python3-wheel\
python3-wheel \
&& pip install --break-system-packages --upgrade setuptools \
&& pip install --break-system-packages -r requirements.txt \
&& apt autoremove -y \

View File

@ -49,17 +49,20 @@ COPY --from=build /app/full /app
WORKDIR /app
ENV PATH="/root/.venv/bin:/root/.local/bin:${PATH}"
# Flag for compatibility with pip
ARG UV_INDEX_STRATEGY="unsafe-best-match"
RUN apt-get update \
&& apt-get install -y \
build-essential \
curl \
git \
python3.13 \
python3.13-dev \
python3-pip \
python3-wheel \
&& update-alternatives --install /usr/bin/python3 python3 /usr/bin/python3.13 100 \
&& pip install --break-system-packages --upgrade setuptools \
&& pip install --break-system-packages -r requirements.txt \
ca-certificates \
&& curl -LsSf https://astral.sh/uv/install.sh | sh \
&& uv python install 3.13 \
&& uv venv --python 3.13 /root/.venv \
&& uv pip install --python /root/.venv/bin/python -r requirements.txt \
&& apt autoremove -y \
&& apt clean -y \
&& rm -rf /tmp/* /var/tmp/* \

View File

@ -25,184 +25,13 @@ permissions:
packages: write
jobs:
push_to_registry:
name: Push Docker image to Docker Hub
runs-on: ${{ matrix.config.runs_on }}
env:
COMMIT_SHA: ${{ github.sha }}
strategy:
fail-fast: false
matrix:
config:
# Multi-stage build
- { tag: "cpu", dockerfile: ".devops/cpu.Dockerfile", platforms: "linux/arm64", full: true, light: true, server: true, free_disk_space: false, runs_on: "ubuntu-24.04" }
- { tag: "cpu", dockerfile: ".devops/cpu.Dockerfile", platforms: "linux/amd64", full: true, light: true, server: true, free_disk_space: false, runs_on: "ubuntu-24.04" }
- { tag: "cuda cuda12", dockerfile: ".devops/cuda.Dockerfile", platforms: "linux/amd64", full: true, light: true, server: true, free_disk_space: true, runs_on: "ubuntu-24.04", cuda_version: "12.4.0", ubuntu_version: "22.04" }
- { tag: "cuda13", dockerfile: ".devops/cuda-new.Dockerfile", platforms: "linux/amd64", full: true, light: true, server: true, free_disk_space: true, runs_on: "ubuntu-24.04", cuda_version: "13.1.0", ubuntu_version: "24.04" }
- { tag: "musa", dockerfile: ".devops/musa.Dockerfile", platforms: "linux/amd64", full: true, light: true, server: true, free_disk_space: true, runs_on: "ubuntu-24.04" }
- { tag: "intel", dockerfile: ".devops/intel.Dockerfile", platforms: "linux/amd64", full: true, light: true, server: true, free_disk_space: true, runs_on: "ubuntu-24.04" }
- { tag: "vulkan", dockerfile: ".devops/vulkan.Dockerfile", platforms: "linux/amd64", full: true, light: true, server: true, free_disk_space: false, runs_on: "ubuntu-24.04" }
- { tag: "s390x", dockerfile: ".devops/s390x.Dockerfile", platforms: "linux/s390x", full: true, light: true, server: true, free_disk_space: false, runs_on: "ubuntu-24.04-s390x" }
- { tag: "rocm", dockerfile: ".devops/rocm.Dockerfile", platforms: "linux/amd64", full: true, light: true, server: true, free_disk_space: true, runs_on: "ubuntu-24.04" }
- { tag: "openvino", dockerfile: ".devops/openvino.Dockerfile", platforms: "linux/amd64", full: true, light: true, server: true, free_disk_space: false, runs_on: "ubuntu-24.04" }
steps:
- name: Check out the repo
uses: actions/checkout@v6
with:
fetch-depth: 0 # preserve git history, so we can determine the build number
- name: Set up QEMU
if: ${{ matrix.config.tag != 's390x' }}
uses: docker/setup-qemu-action@c7c53464625b32c7a7e944ae62b3e17d2b600130 # v3
with:
image: tonistiigi/binfmt:qemu-v10.2.1
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@8d2750c68a42422c14e847fe6c8ac0403b4cbd6f # v3
- name: Log in to Docker Hub
uses: docker/login-action@c94ce9fb468520275223c153574b00df6fe4bcc9 # v3
with:
registry: ghcr.io
username: ${{ github.repository_owner }}
password: ${{ secrets.GITHUB_TOKEN }}
- name: Determine source tag name
id: srctag
uses: ./.github/actions/get-tag-name
env:
BRANCH_NAME: ${{ github.head_ref || github.ref_name }}
- name: Determine image tag name
id: tag
shell: bash
run: |
REPO_OWNER="${GITHUB_REPOSITORY_OWNER@L}" # to lower case
REPO_NAME="${{ github.event.repository.name }}"
PREFIX="ghcr.io/${REPO_OWNER}/${REPO_NAME}:"
# list all tags possible
tags="${{ matrix.config.tag }}"
for tag in $tags; do
if [[ "$tag" == "cpu" ]]; then
TYPE=""
else
TYPE="-$tag"
fi
CACHETAGS="${PREFIX}buildcache${TYPE}"
FULLTAGS="${FULLTAGS:+$FULLTAGS,}${PREFIX}full${TYPE},${PREFIX}full${TYPE}-${{ steps.srctag.outputs.name }}"
LIGHTTAGS="${LIGHTTAGS:+$LIGHTTAGS,}${PREFIX}light${TYPE},${PREFIX}light${TYPE}-${{ steps.srctag.outputs.name }}"
SERVERTAGS="${SERVERTAGS:+$SERVERTAGS,}${PREFIX}server${TYPE},${PREFIX}server${TYPE}-${{ steps.srctag.outputs.name }}"
done
echo "cache_output_tags=$CACHETAGS" >> $GITHUB_OUTPUT
echo "full_output_tags=$FULLTAGS" >> $GITHUB_OUTPUT
echo "light_output_tags=$LIGHTTAGS" >> $GITHUB_OUTPUT
echo "server_output_tags=$SERVERTAGS" >> $GITHUB_OUTPUT
echo "cache_output_tags=$CACHETAGS" # print out for debugging
echo "full_output_tags=$FULLTAGS" # print out for debugging
echo "light_output_tags=$LIGHTTAGS" # print out for debugging
echo "server_output_tags=$SERVERTAGS" # print out for debugging
env:
GITHUB_REPOSITORY_OWNER: '${{ github.repository_owner }}'
- name: Free Disk Space (Ubuntu)
if: ${{ matrix.config.free_disk_space == true }}
uses: ggml-org/free-disk-space@v1.3.1
with:
# this might remove tools that are actually needed,
# if set to "true" but frees about 6 GB
tool-cache: false
# all of these default to true, but feel free to set to
# "false" if necessary for your workflow
android: true
dotnet: true
haskell: true
large-packages: true
docker-images: true
swap-storage: true
- name: Build and push Full Docker image (tagged + versioned)
if: ${{ (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch') && matrix.config.full == true }}
uses: docker/build-push-action@10e90e3645eae34f1e60eeb005ba3a3d33f178e8 # v6
with:
context: .
push: true
platforms: ${{ matrix.config.platforms }}
# tag list is generated from step above
tags: ${{ steps.tag.outputs.full_output_tags }}
file: ${{ matrix.config.dockerfile }}
target: full
provenance: false
build-args: |
${{ matrix.config.ubuntu_version && format('UBUNTU_VERSION={0}', matrix.config.ubuntu_version) || '' }}
${{ matrix.config.cuda_version && format('CUDA_VERSION={0}', matrix.config.cuda_version) || '' }}
# using github experimental cache
#cache-from: type=gha
#cache-to: type=gha,mode=max
# return to this if the experimental github cache is having issues
#cache-to: type=local,dest=/tmp/.buildx-cache
#cache-from: type=local,src=/tmp/.buildx-cache
# using registry cache (no storage limit)
cache-from: type=registry,ref=${{ steps.tag.outputs.cache_output_tags }}
cache-to: type=registry,ref=${{ steps.tag.outputs.cache_output_tags }},mode=max
- name: Build and push Light Docker image (tagged + versioned)
if: ${{ (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch') && matrix.config.light == true }}
uses: docker/build-push-action@10e90e3645eae34f1e60eeb005ba3a3d33f178e8 # v6
with:
context: .
push: true
platforms: ${{ matrix.config.platforms }}
# tag list is generated from step above
tags: ${{ steps.tag.outputs.light_output_tags }}
file: ${{ matrix.config.dockerfile }}
target: light
provenance: false
build-args: |
${{ matrix.config.ubuntu_version && format('UBUNTU_VERSION={0}', matrix.config.ubuntu_version) || '' }}
${{ matrix.config.cuda_version && format('CUDA_VERSION={0}', matrix.config.cuda_version) || '' }}
# using github experimental cache
#cache-from: type=gha
#cache-to: type=gha,mode=max
# return to this if the experimental github cache is having issues
#cache-to: type=local,dest=/tmp/.buildx-cache
#cache-from: type=local,src=/tmp/.buildx-cache
# using registry cache (no storage limit)
cache-from: type=registry,ref=${{ steps.tag.outputs.cache_output_tags }}
cache-to: type=registry,ref=${{ steps.tag.outputs.cache_output_tags }},mode=max
- name: Build and push Server Docker image (tagged + versioned)
if: ${{ (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch') && matrix.config.server == true }}
uses: docker/build-push-action@10e90e3645eae34f1e60eeb005ba3a3d33f178e8 # v6
with:
context: .
push: true
platforms: ${{ matrix.config.platforms }}
# tag list is generated from step above
tags: ${{ steps.tag.outputs.server_output_tags }}
file: ${{ matrix.config.dockerfile }}
target: server
provenance: false
build-args: |
${{ matrix.config.ubuntu_version && format('UBUNTU_VERSION={0}', matrix.config.ubuntu_version) || '' }}
${{ matrix.config.cuda_version && format('CUDA_VERSION={0}', matrix.config.cuda_version) || '' }}
# using github experimental cache
#cache-from: type=gha
#cache-to: type=gha,mode=max
# return to this if the experimental github cache is having issues
#cache-to: type=local,dest=/tmp/.buildx-cache
#cache-from: type=local,src=/tmp/.buildx-cache
# using registry cache (no storage limit)
cache-from: type=registry,ref=${{ steps.tag.outputs.cache_output_tags }}
cache-to: type=registry,ref=${{ steps.tag.outputs.cache_output_tags }},mode=max
create_tag:
name: Create and push git tag
runs-on: ubuntu-22.04
runs-on: ubuntu-slim
permissions:
contents: write
outputs:
source_tag: ${{ steps.srctag.outputs.name }}
steps:
- name: Clone
@ -223,3 +52,391 @@ jobs:
run: |
git tag ${{ steps.srctag.outputs.name }} || exit 0
git push origin ${{ steps.srctag.outputs.name }} || exit 0
prepare_matrices:
name: Prepare Docker matrices
runs-on: ubuntu-24.04
outputs:
build_matrix: ${{ steps.matrices.outputs.build_matrix }}
merge_matrix: ${{ steps.matrices.outputs.merge_matrix }}
steps:
- name: Generate build and merge matrices
id: matrices
shell: bash
run: |
set -euo pipefail
# Keep all build targets in one place and derive merge targets from it.
cat > build-matrix.json <<'JSON'
[
{ "tag": "cpu", "dockerfile": ".devops/cpu.Dockerfile", "platforms": "linux/amd64", "full": true, "light": true, "server": true, "free_disk_space": false, "runs_on": "ubuntu-24.04" },
{ "tag": "cpu", "dockerfile": ".devops/cpu.Dockerfile", "platforms": "linux/arm64", "full": true, "light": true, "server": true, "free_disk_space": false, "runs_on": "ubuntu-24.04-arm" },
{ "tag": "cpu", "dockerfile": ".devops/s390x.Dockerfile", "platforms": "linux/s390x", "full": true, "light": true, "server": true, "free_disk_space": false, "runs_on": "ubuntu-24.04-s390x" },
{ "tag": "cuda cuda12", "dockerfile": ".devops/cuda.Dockerfile", "platforms": "linux/amd64", "full": true, "light": true, "server": true, "free_disk_space": true, "runs_on": "ubuntu-24.04" },
{ "tag": "cuda cuda12", "dockerfile": ".devops/cuda.Dockerfile", "platforms": "linux/arm64", "full": true, "light": true, "server": true, "free_disk_space": true, "runs_on": "ubuntu-24.04-arm" },
{ "tag": "cuda13", "dockerfile": ".devops/cuda-new.Dockerfile", "platforms": "linux/amd64", "full": true, "light": true, "server": true, "free_disk_space": true, "runs_on": "ubuntu-24.04" },
{ "tag": "cuda13", "dockerfile": ".devops/cuda-new.Dockerfile", "platforms": "linux/arm64", "full": true, "light": true, "server": true, "free_disk_space": true, "runs_on": "ubuntu-24.04-arm" },
{ "tag": "musa", "dockerfile": ".devops/musa.Dockerfile", "platforms": "linux/amd64", "full": true, "light": true, "server": true, "free_disk_space": true, "runs_on": "ubuntu-24.04" },
{ "tag": "intel", "dockerfile": ".devops/intel.Dockerfile", "platforms": "linux/amd64", "full": true, "light": true, "server": true, "free_disk_space": true, "runs_on": "ubuntu-24.04" },
{ "tag": "vulkan", "dockerfile": ".devops/vulkan.Dockerfile", "platforms": "linux/amd64", "full": true, "light": true, "server": true, "free_disk_space": false, "runs_on": "ubuntu-24.04" },
{ "tag": "vulkan", "dockerfile": ".devops/vulkan.Dockerfile", "platforms": "linux/arm64", "full": true, "light": true, "server": true, "free_disk_space": false, "runs_on": "ubuntu-24.04-arm" },
{ "tag": "rocm", "dockerfile": ".devops/rocm.Dockerfile", "platforms": "linux/amd64", "full": true, "light": true, "server": true, "free_disk_space": true, "runs_on": "ubuntu-24.04" },
{ "tag": "openvino", "dockerfile": ".devops/openvino.Dockerfile", "platforms": "linux/amd64", "full": true, "light": true, "server": true, "free_disk_space": false, "runs_on": "ubuntu-24.04" }
]
JSON
BUILD_MATRIX="$(jq -c . build-matrix.json)"
MERGE_MATRIX="$(jq -c '
reduce .[] as $entry ({}; .[$entry.tag] |= (
. // {
tag: $entry.tag,
arches: [],
full: false,
light: false,
server: false
}
| .full = (.full or ($entry.full // false))
| .light = (.light or ($entry.light // false))
| .server = (.server or ($entry.server // false))
| .arches += [($entry.platforms | sub("^linux/"; ""))]
))
# Backward compatibility: s390x tags are aliases of cpu for the linux/s390x platform.
| if (has("cpu") and (((.cpu.arches // []) | index("s390x")) != null)) then
. + {
s390x: {
tag: "s390x",
arches: ["s390x"],
full: .cpu.full,
light: .cpu.light,
server: .cpu.server
}
}
else
.
end
| [.[] | .arches = (.arches | unique | sort | join(" "))]
' build-matrix.json)"
echo "build_matrix=$BUILD_MATRIX" >> "$GITHUB_OUTPUT"
echo "merge_matrix=$MERGE_MATRIX" >> "$GITHUB_OUTPUT"
push_to_registry:
name: Push Docker image to Docker Registry
needs: [prepare_matrices, create_tag]
runs-on: ${{ matrix.config.runs_on }}
strategy:
fail-fast: false
matrix:
config: ${{ fromJSON(needs.prepare_matrices.outputs.build_matrix) }}
steps:
- name: Check out the repo
uses: actions/checkout@v6
with:
fetch-depth: 0
ref: ${{ needs.create_tag.outputs.source_tag }}
- name: Set up QEMU
if: ${{ contains(matrix.config.platforms, 'linux/amd64') }}
uses: docker/setup-qemu-action@ce360397dd3f832beb865e1373c09c0e9f86d70a # v4
with:
image: tonistiigi/binfmt:qemu-v10.2.1
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@4d04d5d9486b7bd6fa91e7baf45bbb4f8b9deedd # v4
- name: Log in to Docker Registry
uses: docker/login-action@b45d80f862d83dbcd57f89517bcf500b2ab88fb2 # v4
with:
registry: ghcr.io
username: ${{ github.repository_owner }}
password: ${{ secrets.GITHUB_TOKEN }}
- name: Determine image metadata
id: meta
shell: bash
run: |
set -euo pipefail
REPO_OWNER="${GITHUB_REPOSITORY_OWNER@L}" # to lower case
REPO_NAME="${{ github.event.repository.name }}"
IMAGE_REPO="ghcr.io/${REPO_OWNER}/${REPO_NAME}"
PREFIX="${IMAGE_REPO}:"
PLATFORM="${{ matrix.config.platforms }}"
ARCH_SUFFIX="${PLATFORM#linux/}"
# list all tags possible
tags="${{ matrix.config.tag }}"
for tag in $tags; do
if [[ "$tag" == "cpu" ]]; then
TYPE=""
else
TYPE="-$tag"
fi
CACHETAG="${PREFIX}buildcache${TYPE}-${ARCH_SUFFIX}"
done
SAFE_TAGS="$(echo "$tags" | tr ' ' '_')"
echo "image_repo=$IMAGE_REPO" >> $GITHUB_OUTPUT
echo "arch_suffix=$ARCH_SUFFIX" >> $GITHUB_OUTPUT
echo "cache_output_tag=$CACHETAG" >> $GITHUB_OUTPUT
echo "digest_artifact_suffix=${SAFE_TAGS}-${ARCH_SUFFIX}" >> $GITHUB_OUTPUT
echo "cache_output_tag=$CACHETAG" # print out for debugging
env:
GITHUB_REPOSITORY_OWNER: '${{ github.repository_owner }}'
- name: Free Disk Space (Ubuntu)
if: ${{ matrix.config.free_disk_space == true }}
uses: ggml-org/free-disk-space@v1.3.1
with:
# this might remove tools that are actually needed,
# if set to "true" but frees about 6 GB
tool-cache: false
# all of these default to true, but feel free to set to
# "false" if necessary for your workflow
android: true
dotnet: true
haskell: true
large-packages: true
docker-images: true
swap-storage: true
- name: Build and push Full Docker image by digest
id: build_full
if: ${{ (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch') && matrix.config.full == true }}
uses: docker/build-push-action@d08e5c354a6adb9ed34480a06d141179aa583294 # v7
with:
context: .
platforms: ${{ matrix.config.platforms }}
outputs: type=image,name=${{ steps.meta.outputs.image_repo }},push-by-digest=true,name-canonical=true,push=true
file: ${{ matrix.config.dockerfile }}
target: full
provenance: false
build-args: |
${{ matrix.config.ubuntu_version && format('UBUNTU_VERSION={0}', matrix.config.ubuntu_version) || '' }}
${{ matrix.config.cuda_version && format('CUDA_VERSION={0}', matrix.config.cuda_version) || '' }}
# using github experimental cache
#cache-from: type=gha
#cache-to: type=gha,mode=max
# return to this if the experimental github cache is having issues
#cache-to: type=local,dest=/tmp/.buildx-cache
#cache-from: type=local,src=/tmp/.buildx-cache
# using registry cache (no storage limit)
cache-from: type=registry,ref=${{ steps.meta.outputs.cache_output_tag }}
cache-to: type=registry,ref=${{ steps.meta.outputs.cache_output_tag }},mode=max
- name: Build and push Light Docker image by digest
id: build_light
if: ${{ (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch') && matrix.config.light == true }}
uses: docker/build-push-action@d08e5c354a6adb9ed34480a06d141179aa583294 # v7
with:
context: .
platforms: ${{ matrix.config.platforms }}
outputs: type=image,name=${{ steps.meta.outputs.image_repo }},push-by-digest=true,name-canonical=true,push=true
file: ${{ matrix.config.dockerfile }}
target: light
provenance: false
build-args: |
${{ matrix.config.ubuntu_version && format('UBUNTU_VERSION={0}', matrix.config.ubuntu_version) || '' }}
${{ matrix.config.cuda_version && format('CUDA_VERSION={0}', matrix.config.cuda_version) || '' }}
# using github experimental cache
#cache-from: type=gha
#cache-to: type=gha,mode=max
# return to this if the experimental github cache is having issues
#cache-to: type=local,dest=/tmp/.buildx-cache
#cache-from: type=local,src=/tmp/.buildx-cache
# using registry cache (no storage limit)
cache-from: type=registry,ref=${{ steps.meta.outputs.cache_output_tag }}
cache-to: type=registry,ref=${{ steps.meta.outputs.cache_output_tag }},mode=max
- name: Build and push Server Docker image by digest
id: build_server
if: ${{ (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch') && matrix.config.server == true }}
uses: docker/build-push-action@d08e5c354a6adb9ed34480a06d141179aa583294 # v7
with:
context: .
platforms: ${{ matrix.config.platforms }}
outputs: type=image,name=${{ steps.meta.outputs.image_repo }},push-by-digest=true,name-canonical=true,push=true
file: ${{ matrix.config.dockerfile }}
target: server
provenance: false
build-args: |
${{ matrix.config.ubuntu_version && format('UBUNTU_VERSION={0}', matrix.config.ubuntu_version) || '' }}
${{ matrix.config.cuda_version && format('CUDA_VERSION={0}', matrix.config.cuda_version) || '' }}
# using github experimental cache
#cache-from: type=gha
#cache-to: type=gha,mode=max
# return to this if the experimental github cache is having issues
#cache-to: type=local,dest=/tmp/.buildx-cache
#cache-from: type=local,src=/tmp/.buildx-cache
# using registry cache (no storage limit)
cache-from: type=registry,ref=${{ steps.meta.outputs.cache_output_tag }}
cache-to: type=registry,ref=${{ steps.meta.outputs.cache_output_tag }},mode=max
- name: Export digest metadata
shell: bash
run: |
set -euo pipefail
TAGS="${{ matrix.config.tag }}"
ARCH_SUFFIX="${{ steps.meta.outputs.arch_suffix }}"
DIGEST_FILE="/tmp/digests/${{ steps.meta.outputs.digest_artifact_suffix }}.tsv"
mkdir -p /tmp/digests
add_digest_rows() {
local image_type="$1"
local digest="$2"
if [[ -z "$digest" ]]; then
echo "Missing digest for image_type=${image_type}" >&2
exit 1
fi
for tag in $TAGS; do
printf '%s\t%s\t%s\t%s\n' "$tag" "$ARCH_SUFFIX" "$image_type" "$digest" >> "$DIGEST_FILE"
done
}
if [[ "${{ matrix.config.full }}" == "true" ]]; then
add_digest_rows "full" "${{ steps.build_full.outputs.digest }}"
fi
if [[ "${{ matrix.config.light }}" == "true" ]]; then
add_digest_rows "light" "${{ steps.build_light.outputs.digest }}"
fi
if [[ "${{ matrix.config.server }}" == "true" ]]; then
add_digest_rows "server" "${{ steps.build_server.outputs.digest }}"
fi
- name: Upload digest metadata
uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f # v7
with:
name: digests-${{ steps.meta.outputs.digest_artifact_suffix }}
path: /tmp/digests/${{ steps.meta.outputs.digest_artifact_suffix }}.tsv
if-no-files-found: error
merge_arch_tags:
name: Create shared tags from digests
needs: [prepare_matrices, push_to_registry, create_tag]
runs-on: ubuntu-24.04
strategy:
fail-fast: false
matrix:
config: ${{ fromJSON(needs.prepare_matrices.outputs.merge_matrix) }}
steps:
- name: Check out the repo
uses: actions/checkout@v6
with:
fetch-depth: 0
- name: Download digest metadata
uses: actions/download-artifact@3e5f45b2cfb9172054b4087a40e8e0b5a5461e7c # v8
with:
pattern: digests-*
path: /tmp/digests
merge-multiple: true
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@4d04d5d9486b7bd6fa91e7baf45bbb4f8b9deedd # v4
- name: Log in to Docker Registry
uses: docker/login-action@b45d80f862d83dbcd57f89517bcf500b2ab88fb2 # v4
with:
registry: ghcr.io
username: ${{ github.repository_owner }}
password: ${{ secrets.GITHUB_TOKEN }}
- name: Create tags from digests
shell: bash
run: |
set -euo pipefail
REPO_OWNER="${GITHUB_REPOSITORY_OWNER@L}" # to lower case
REPO_NAME="${{ github.event.repository.name }}"
IMAGE_REPO="ghcr.io/${REPO_OWNER}/${REPO_NAME}"
PREFIX="${IMAGE_REPO}:"
SRC_TAG="${{ needs.create_tag.outputs.source_tag }}"
TAGS="${{ matrix.config.tag }}"
ARCHES="${{ matrix.config.arches }}"
DIGEST_GLOB="/tmp/digests/*.tsv"
if ! ls ${DIGEST_GLOB} >/dev/null 2>&1; then
echo "No digest metadata found in /tmp/digests" >&2
exit 1
fi
if [[ -z "$SRC_TAG" ]]; then
echo "Missing source tag from create_tag" >&2
exit 1
fi
find_digest() {
local tag_name="$1"
local arch="$2"
local image_type="$3"
local digest
digest="$(awk -F '\t' -v t="$tag_name" -v a="$arch" -v i="$image_type" '$1 == t && $2 == a && $3 == i { print $4; exit }' ${DIGEST_GLOB})"
# Backward compatibility: s390x tags are aliases of cpu for the linux/s390x platform.
if [[ -z "$digest" && "$tag_name" == "s390x" && "$arch" == "s390x" ]]; then
digest="$(awk -F '\t' -v t="cpu" -v a="$arch" -v i="$image_type" '$1 == t && $2 == a && $3 == i { print $4; exit }' ${DIGEST_GLOB})"
fi
if [[ -z "$digest" ]]; then
echo "Missing digest for tag=${tag_name} arch=${arch} image_type=${image_type}" >&2
exit 1
fi
echo "$digest"
}
create_manifest_tags() {
local image_type="$1"
local tag_name="$2"
local suffix="$3"
local merged_tag="${PREFIX}${image_type}${suffix}"
local merged_versioned_tag="${merged_tag}-${SRC_TAG}"
local refs=()
for arch in $ARCHES; do
local digest
digest="$(find_digest "$tag_name" "$arch" "$image_type")"
refs+=("${IMAGE_REPO}@${digest}")
done
echo "Creating ${merged_tag} from ${refs[*]}"
docker buildx imagetools create --tag "${merged_tag}" "${refs[@]}"
echo "Creating ${merged_versioned_tag} from ${refs[*]}"
docker buildx imagetools create --tag "${merged_versioned_tag}" "${refs[@]}"
}
for tag in $TAGS; do
if [[ "$tag" == "cpu" ]]; then
TYPE=""
else
TYPE="-$tag"
fi
if [[ "${{ matrix.config.full }}" == "true" ]]; then
create_manifest_tags "full" "$tag" "$TYPE"
fi
if [[ "${{ matrix.config.light }}" == "true" ]]; then
create_manifest_tags "light" "$tag" "$TYPE"
fi
if [[ "${{ matrix.config.server }}" == "true" ]]; then
create_manifest_tags "server" "$tag" "$TYPE"
fi
done
env:
GITHUB_REPOSITORY_OWNER: '${{ github.repository_owner }}'

View File

@ -13,24 +13,30 @@ We have three Docker images available for this project:
Additionally, there the following images, similar to the above:
- `ghcr.io/ggml-org/llama.cpp:full-cuda`: Same as `full` but compiled with CUDA support. (platforms: `linux/amd64`)
- `ghcr.io/ggml-org/llama.cpp:light-cuda`: Same as `light` but compiled with CUDA support. (platforms: `linux/amd64`)
- `ghcr.io/ggml-org/llama.cpp:server-cuda`: Same as `server` but compiled with CUDA support. (platforms: `linux/amd64`)
- `ghcr.io/ggml-org/llama.cpp:full-rocm`: Same as `full` but compiled with ROCm support. (platforms: `linux/amd64`, `linux/arm64`)
- `ghcr.io/ggml-org/llama.cpp:light-rocm`: Same as `light` but compiled with ROCm support. (platforms: `linux/amd64`, `linux/arm64`)
- `ghcr.io/ggml-org/llama.cpp:server-rocm`: Same as `server` but compiled with ROCm support. (platforms: `linux/amd64`, `linux/arm64`)
- `ghcr.io/ggml-org/llama.cpp:full-cuda`: Same as `full` but compiled with CUDA 12 support. (platforms: `linux/amd64`, `linux/arm64`)
- `ghcr.io/ggml-org/llama.cpp:full-cuda13`: Same as `full` but compiled with CUDA 13 support. (platforms: `linux/amd64`, `linux/arm64`)
- `ghcr.io/ggml-org/llama.cpp:light-cuda`: Same as `light` but compiled with CUDA 12 support. (platforms: `linux/amd64`, `linux/arm64`)
- `ghcr.io/ggml-org/llama.cpp:light-cuda13`: Same as `light` but compiled with CUDA 13 support. (platforms: `linux/amd64`, `linux/arm64`)
- `ghcr.io/ggml-org/llama.cpp:server-cuda`: Same as `server` but compiled with CUDA 12 support. (platforms: `linux/amd64`, `linux/arm64`)
- `ghcr.io/ggml-org/llama.cpp:server-cuda13`: Same as `server` but compiled with CUDA 13 support. (platforms: `linux/amd64`, `linux/arm64`)
- `ghcr.io/ggml-org/llama.cpp:full-rocm`: Same as `full` but compiled with ROCm support. (platforms: `linux/amd64`)
- `ghcr.io/ggml-org/llama.cpp:light-rocm`: Same as `light` but compiled with ROCm support. (platforms: `linux/amd64`)
- `ghcr.io/ggml-org/llama.cpp:server-rocm`: Same as `server` but compiled with ROCm support. (platforms: `linux/amd64`)
- `ghcr.io/ggml-org/llama.cpp:full-musa`: Same as `full` but compiled with MUSA support. (platforms: `linux/amd64`)
- `ghcr.io/ggml-org/llama.cpp:light-musa`: Same as `light` but compiled with MUSA support. (platforms: `linux/amd64`)
- `ghcr.io/ggml-org/llama.cpp:server-musa`: Same as `server` but compiled with MUSA support. (platforms: `linux/amd64`)
- `ghcr.io/ggml-org/llama.cpp:full-intel`: Same as `full` but compiled with SYCL support. (platforms: `linux/amd64`)
- `ghcr.io/ggml-org/llama.cpp:light-intel`: Same as `light` but compiled with SYCL support. (platforms: `linux/amd64`)
- `ghcr.io/ggml-org/llama.cpp:server-intel`: Same as `server` but compiled with SYCL support. (platforms: `linux/amd64`)
- `ghcr.io/ggml-org/llama.cpp:full-vulkan`: Same as `full` but compiled with Vulkan support. (platforms: `linux/amd64`)
- `ghcr.io/ggml-org/llama.cpp:light-vulkan`: Same as `light` but compiled with Vulkan support. (platforms: `linux/amd64`)
- `ghcr.io/ggml-org/llama.cpp:server-vulkan`: Same as `server` but compiled with Vulkan support. (platforms: `linux/amd64`)
- `ghcr.io/ggml-org/llama.cpp:full-vulkan`: Same as `full` but compiled with Vulkan support. (platforms: `linux/amd64`, `linux/arm64`)
- `ghcr.io/ggml-org/llama.cpp:light-vulkan`: Same as `light` but compiled with Vulkan support. (platforms: `linux/amd64`, `linux/arm64`)
- `ghcr.io/ggml-org/llama.cpp:server-vulkan`: Same as `server` but compiled with Vulkan support. (platforms: `linux/amd64`, `linux/arm64`)
- `ghcr.io/ggml-org/llama.cpp:full-openvino`: Same as `full` but compiled with OpenVino support. (platforms: `linux/amd64`)
- `ghcr.io/ggml-org/llama.cpp:light-openvino`: Same as `light` but compiled with OpenVino support. (platforms: `linux/amd64`)
- `ghcr.io/ggml-org/llama.cpp:server-openvino`: Same as `server` but compiled with OpenVino support. (platforms: `linux/amd64`)
- `ghcr.io/ggml-org/llama.cpp:full-s390x`: Identical to `full`, an alias for the `s390x` platform. (platforms: `linux/s390x`)
- `ghcr.io/ggml-org/llama.cpp:light-s390x`: Identical to `light`, an alias for the `s390x` platform. (platforms: `linux/s390x`)
- `ghcr.io/ggml-org/llama.cpp:server-s390x`: Identical to `server`, an alias for the `s390x` platform. (platforms: `linux/s390x`)
The GPU enabled images are not currently tested by CI beyond being built. They are not built with any variation from the ones in the Dockerfiles defined in [.devops/](../.devops/) and the GitHub Action defined in [.github/workflows/docker.yml](../.github/workflows/docker.yml). If you need different settings (for example, a different CUDA, ROCm or MUSA library, you'll need to build the images locally for now).
@ -82,7 +88,7 @@ You may want to pass in some different `ARGS`, depending on the CUDA environment
The defaults are:
- `CUDA_VERSION` set to `12.4.0`
- `CUDA_VERSION` set to `12.8.1`
- `CUDA_DOCKER_ARCH` set to the cmake build default, which includes all the supported architectures
The resulting images, are essentially the same as the non-CUDA images: