wylab/llama-swap-rocm
1
0
Fork 0
forked from wylab/llama-swap
Code Pull Requests Actions Packages Activity

ci: adapt build for Gitea registry, drop ghcr.io dependency

Browse Source 
- Login to git.wylab.me instead of ghcr.io
- Use Gitea-hosted llama.cpp-rocm base image instead of ghcr.io
- Rewrite fetch_llama_tag to use anonymous OCI registry API
- Add LS_UPSTREAM for release binary fetches on forks
- Add REGISTRY and BASE_TAG overrides for self-hosted builds
- Only build rocm platform

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
This commit is contained in:
claude
2026-04-12 19:39:57 +02:00
parent a9d840ffd7
commit 980a4e6baa
2 changed files with 66 additions and 107 deletions
Download Patch File Download Diff File Expand all files Collapse all files
.github/workflows/containers.yml
+7 -46
View File
@@ -1,12 +1,6 @@
name: Build Containers
on:
# time has no specific meaning, trying to time it after
# the llama.cpp daily packages are published
# https://github.com/ggml-org/llama.cpp/blob/master/.github/workflows/docker.yml
schedule:
- cron: "37 5 * * *"
# Allows manual triggering of the workflow
workflow_dispatch:
@@ -17,57 +11,24 @@ on:
- 'docker/build-container.sh'
- 'docker/*.Containerfile'
# grant permissions on GITHUB_TOKEN to publish packages
# ref: https://docs.github.com/en/packages/managing-github-packages-using-github-actions-workflows/publishing-and-installing-a-package-with-github-actions#publishing-a-package-using-an-action
permissions:
contents: read
packages: write
id-token: write
jobs:
build-and-push:
runs-on: ubuntu-latest
strategy:
matrix:
platform: [intel, cuda, cuda13, vulkan, cpu, musa, rocm]
platform: [rocm]
fail-fast: false
steps:
- name: Checkout code
uses: actions/checkout@v4
- name: Free up disk space
if: matrix.platform == 'rocm'
- name: Log in to Gitea Container Registry
run: |
echo "Before cleanup:"
df -h
sudo rm -rf /usr/share/dotnet
sudo rm -rf /usr/local/lib/android
sudo rm -rf /opt/ghc
sudo rm -rf /opt/hostedtoolcache/CodeQL
sudo docker system prune -af
echo "After cleanup:"
df -h
- name: Log in to GitHub Container Registry
uses: docker/login-action@v2
with:
registry: ghcr.io
username: ${{ github.actor }}
password: ${{ secrets.GITHUB_TOKEN }}
echo "${{ secrets.GITEA_TOKEN || github.token }}" | docker login git.wylab.me -u ${{ github.actor }} --password-stdin
- name: Run build-container
env:
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
run: ./docker/build-container.sh ${{ matrix.platform }} ${{ github.event_name != 'push' }}
# note make sure mostlygeek/llama-swap has admin rights to the llama-swap package
# see: https://github.com/actions/delete-package-versions/issues/74
delete-untagged-containers:
needs: build-and-push
runs-on: ubuntu-latest
steps:
- uses: actions/delete-package-versions@v5
with:
package-name: 'llama-swap'
package-type: 'container'
delete-only-untagged-versions: 'true'
REGISTRY: git.wylab.me
BASE_LLAMACPP_IMAGE: git.wylab.me/wylab/llama.cpp-rocm
BASE_TAG: server-rocm
run: ./docker/build-container.sh ${{ matrix.platform }} true
docker/build-container.sh
+59 -61
View File
@@ -35,100 +35,98 @@ if [[ ! " ${ALLOWED_ARCHS[@]} " =~ " ${ARCH} " ]]; then
exit 1
fi
# Check if GITHUB_TOKEN is set and not empty
if [[ -z "${GITHUB_TOKEN:-}" ]]; then
log_info "Error: GITHUB_TOKEN is not set or is empty."
exit 1
fi
# Set llama.cpp base image, customizable using the BASE_LLAMACPP_IMAGE environment
# variable, this permits testing with forked llama.cpp repositories
BASE_IMAGE=${BASE_LLAMACPP_IMAGE:-ghcr.io/ggml-org/llama.cpp}
BASE_IMAGE=${BASE_LLAMACPP_IMAGE:-git.wylab.me/wylab/llama.cpp-rocm}
SD_IMAGE=${BASE_SDCPP_IMAGE:-ghcr.io/leejet/stable-diffusion.cpp}
# Set llama-swap repository, automatically uses GITHUB_REPOSITORY variable
# to enable easy container builds on forked repos
LS_REPO=${GITHUB_REPOSITORY:-mostlygeek/llama-swap}
# Upstream repo for fetching release binaries (always mostlygeek, even on forks)
LS_UPSTREAM=${LS_UPSTREAM:-mostlygeek/llama-swap}
# the most recent llama-swap tag
# have to strip out the 'v' due to .tar.gz file naming
LS_VER=$(curl -s https://api.github.com/repos/${LS_REPO}/releases/latest | jq -r .tag_name | sed 's/v//')
LS_VER=$(curl -s https://api.github.com/repos/${LS_UPSTREAM}/releases/latest | jq -r .tag_name | sed 's/v//')
# Fetches the most recent llama.cpp tag matching the given prefix
# Handles pagination to search beyond the first 100 results
# Uses the OCI registry API (anonymous, no GITHUB_TOKEN needed)
# $1 - tag_prefix (e.g., "server" or "server-vulkan")
# Returns: the version number extracted from the tag
fetch_llama_tag() {
local tag_prefix=$1
local page=1
local per_page=100
# Get anonymous bearer token for ghcr.io
local token=$(curl -s "https://ghcr.io/token?scope=repository:ggml-org/llama.cpp:pull" | jq -r .token)
if [ -z "$token" ] || [ "$token" = "null" ]; then
log_info "Failed to get ghcr.io anonymous token"
return 1
fi
# Paginate through OCI tag list
local all_tags=""
local last=""
while true; do
log_debug "Fetching page $page for tag prefix: $tag_prefix"
local url="https://ghcr.io/v2/ggml-org/llama.cpp/tags/list?n=1000"
if [ -n "$last" ]; then url="${url}&last=${last}"; fi
local response=$(curl -s -H "Authorization: Bearer $GITHUB_TOKEN" \
"https://api.github.com/users/ggml-org/packages/container/llama.cpp/versions?per_page=${per_page}&page=${page}")
local batch=$(curl -s -H "Authorization: Bearer $token" "$url" | jq -r '.tags[]' 2>/dev/null)
if [ -z "$batch" ]; then break; fi
# Check for API errors
if echo "$response" | jq -e '.message' > /dev/null 2>&1; then
local error_msg=$(echo "$response" | jq -r '.message')
log_info "GitHub API error: $error_msg"
return 1
fi
# Check if response is empty array (no more pages)
if [ "$(echo "$response" | jq 'length')" -eq 0 ]; then
log_debug "No more pages (empty response)"
return 1
fi
# Extract matching tag from this page
local found_tag=$(echo "$response" | jq -r \
".[] | select(.metadata.container.tags[]? | startswith(\"$tag_prefix\")) | .metadata.container.tags[] | select(startswith(\"$tag_prefix\"))" \
| sort -r | head -n1)
if [ -n "$found_tag" ]; then
log_debug "Found tag: $found_tag on page $page"
echo "$found_tag" | awk -F '-' '{print $NF}'
return 0
fi
page=$((page + 1))
# Safety limit to prevent infinite loops
if [ $page -gt 50 ]; then
log_info "Reached pagination safety limit (50 pages)"
return 1
fi
all_tags="${all_tags}${batch}"$'\n'
last=$(echo "$batch" | tail -1)
local count=$(echo "$batch" | wc -l)
if [ "$count" -lt 1000 ]; then break; fi
done
# Find the latest tag matching the prefix
local found_tag=$(echo "$all_tags" | grep "^${tag_prefix}-" | sort -V | tail -1)
if [ -n "$found_tag" ]; then
log_debug "Found tag: $found_tag"
echo "$found_tag" | awk -F '-' '{print $NF}'
return 0
fi
log_info "No tag found matching prefix: $tag_prefix"
return 1
}
if [ "$ARCH" == "cpu" ]; then
LCPP_TAG=$(fetch_llama_tag "server")
BASE_TAG=server-${LCPP_TAG}
# BASE_TAG can be overridden to skip tag discovery (e.g. BASE_TAG=server-rocm)
if [[ -z "${BASE_TAG:-}" ]]; then
if [ "$ARCH" == "cpu" ]; then
LCPP_TAG=$(fetch_llama_tag "server")
BASE_TAG=server-${LCPP_TAG}
else
LCPP_TAG=$(fetch_llama_tag "server-${ARCH}")
BASE_TAG=server-${ARCH}-${LCPP_TAG}
fi
# Abort if LCPP_TAG is empty.
if [[ -z "$LCPP_TAG" ]]; then
log_info "Abort: Could not find llama-server container for arch: $ARCH"
exit 1
fi
log_info "LCPP_TAG: $LCPP_TAG"
else
LCPP_TAG=$(fetch_llama_tag "server-${ARCH}")
BASE_TAG=server-${ARCH}-${LCPP_TAG}
LCPP_TAG=${BASE_TAG##*-}
log_info "Using override BASE_TAG: $BASE_TAG (LCPP_TAG: $LCPP_TAG)"
fi
SD_TAG=master-${ARCH}
# Abort if LCPP_TAG is empty.
if [[ -z "$LCPP_TAG" ]]; then
log_info "Abort: Could not find llama-server container for arch: $ARCH"
exit 1
else
log_info "LCPP_TAG: $LCPP_TAG"
fi
if [[ ! -z "$DEBUG_ABORT_BUILD" ]]; then
log_info "Abort: DEBUG_ABORT_BUILD set"
exit 0
fi
# Registry defaults to ghcr.io, can be overridden via REGISTRY env var
REGISTRY=${REGISTRY:-ghcr.io}
for CONTAINER_TYPE in non-root root; do
CONTAINER_TAG="ghcr.io/${LS_REPO}:v${LS_VER}-${ARCH}-${LCPP_TAG}"
CONTAINER_LATEST="ghcr.io/${LS_REPO}:${ARCH}"
CONTAINER_TAG="${REGISTRY}/${LS_REPO}:v${LS_VER}-${ARCH}-${LCPP_TAG}"
CONTAINER_LATEST="${REGISTRY}/${LS_REPO}:${ARCH}"
USER_UID=0
USER_GID=0
USER_HOME=/root
@@ -143,7 +141,7 @@ for CONTAINER_TYPE in non-root root; do
log_info "Building $CONTAINER_TYPE $CONTAINER_TAG $LS_VER"
docker build --provenance=false -f llama-swap.Containerfile --build-arg BASE_TAG=${BASE_TAG} --build-arg LS_VER=${LS_VER} --build-arg UID=${USER_UID} \
--build-arg LS_REPO=${LS_REPO} --build-arg GID=${USER_GID} --build-arg USER_HOME=${USER_HOME} -t ${CONTAINER_TAG} -t ${CONTAINER_LATEST} \
--build-arg LS_REPO=${LS_UPSTREAM} --build-arg GID=${USER_GID} --build-arg USER_HOME=${USER_HOME} -t ${CONTAINER_TAG} -t ${CONTAINER_LATEST} \
--build-arg BASE_IMAGE=${BASE_IMAGE} .
# For architectures with stable-diffusion.cpp support, layer sd-server on top