Skip to content

fix: Propagate version tag to WebUI asset download in self-hosted CI #23051

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
allozaur merged 8 commits into from
May 14, 2026
Merged

fix: Propagate version tag to WebUI asset download in self-hosted CI #23051

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
8 commits
Select commit Hold shift + click to select a range
47cf1f5
fix: Propagate version tag to WebUI asset download in self-hosted CI
allozaur May 14, 2026
69d3634
refactor: Apply suggestions from @CISC
allozaur May 14, 2026
a7f9a81
fix: Skip npm build when Node.js is not installed
allozaur May 14, 2026
75f2129
fix: Use + separator for ASSETS list to fix Windows build
allozaur May 14, 2026
172840e
fix: Validate HF_WEBUI_VERSION environment variable with regex
allozaur May 14, 2026
17a822c
fix: Remove 'latest' fallback for HF_WEBUI_VERSION
allozaur May 14, 2026
1285374
fix: Demote checksum verification failure to warning instead of hard …
allozaur May 14, 2026
a0fa681
fix: End line character
allozaur May 14, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
40 changes: 40 additions & 0 deletions .github/workflows/build-self-hosted.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -55,7 +55,22 @@ env:
LLAMA_LOG_TIMESTAMPS: 1

jobs:
determine-tag:
name: Determine tag name
runs-on: ubuntu-latest
Comment thread
allozaur marked this conversation as resolved.
Outdated
outputs:
tag_name: ${{ steps.tag.outputs.name }}
steps:
- name: Clone
uses: actions/checkout@v6
with:
fetch-depth: 0
- name: Determine tag name
id: tag
uses: ./.github/actions/get-tag-name

Comment on lines +58 to +71

@ggerganov ggerganov May 24, 2026

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Do we actually need a tag for these jobs? I don't think they require the webui at all.

@CISC CISC May 24, 2026

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I've wondered myself, particularly because the webgpu runner does not depend on it.

@ggerganov ggerganov May 24, 2026

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yes, I'm running a test without this job: https://github.com/ggml-org/llama.cpp/actions/runs/26364823846

If it passes will open a PR.

ggml-ci-nvidia-cuda:
needs: determine-tag
runs-on: [self-hosted, Linux, NVIDIA]

steps:
Expand All @@ -65,11 +80,14 @@ jobs:

- name: Test
id: ggml-ci
env:
HF_WEBUI_VERSION: ${{ needs.determine-tag.outputs.tag_name || 'latest' }}
Comment thread
allozaur marked this conversation as resolved.
Outdated
run: |
nvidia-smi
GG_BUILD_CUDA=1 bash ./ci/run.sh ~/results/llama.cpp /mnt/llama.cpp

ggml-ci-nvidia-vulkan-cm:
needs: determine-tag
runs-on: [self-hosted, Linux, NVIDIA]

steps:
Expand All @@ -79,11 +97,14 @@ jobs:

- name: Test
id: ggml-ci
env:
HF_WEBUI_VERSION: ${{ needs.determine-tag.outputs.tag_name || 'latest' }}
run: |
vulkaninfo --summary
GG_BUILD_VULKAN=1 GGML_VK_DISABLE_COOPMAT2=1 bash ./ci/run.sh ~/results/llama.cpp /mnt/llama.cpp

ggml-ci-nvidia-vulkan-cm2:
needs: determine-tag
runs-on: [self-hosted, Linux, NVIDIA, COOPMAT2]

steps:
Expand All @@ -93,6 +114,8 @@ jobs:

- name: Test
id: ggml-ci
env:
HF_WEBUI_VERSION: ${{ needs.determine-tag.outputs.tag_name || 'latest' }}
run: |
vulkaninfo --summary
GG_BUILD_VULKAN=1 bash ./ci/run.sh ~/results/llama.cpp /mnt/llama.cpp
Expand Down Expand Up @@ -172,6 +195,7 @@ jobs:
# GG_BUILD_ROCM=1 GG_BUILD_AMDGPU_TARGETS="gfx1101" bash ./ci/run.sh ~/results/llama.cpp /mnt/llama.cpp

ggml-ci-mac-metal:
needs: determine-tag
runs-on: [self-hosted, macOS, ARM64]

steps:
Expand All @@ -181,10 +205,13 @@ jobs:

- name: Test
id: ggml-ci
env:
HF_WEBUI_VERSION: ${{ needs.determine-tag.outputs.tag_name || 'latest' }}
run: |
GG_BUILD_METAL=1 bash ./ci/run.sh ~/results/llama.cpp ~/mnt/llama.cpp

ggml-ci-mac-webgpu:
needs: determine-tag
runs-on: [self-hosted, macOS, ARM64]

steps:
Expand All @@ -207,11 +234,14 @@ jobs:

- name: Test
id: ggml-ci
env:
HF_WEBUI_VERSION: ${{ needs.determine-tag.outputs.tag_name || 'latest' }}
run: |
GG_BUILD_WEBGPU=1 GG_BUILD_WEBGPU_DAWN_PREFIX="$GITHUB_WORKSPACE/dawn" \
bash ./ci/run.sh ~/results/llama.cpp ~/mnt/llama.cpp

ggml-ci-mac-vulkan:
needs: determine-tag
runs-on: [self-hosted, macOS, ARM64]

steps:
Expand All @@ -221,11 +251,14 @@ jobs:

- name: Test
id: ggml-ci
env:
HF_WEBUI_VERSION: ${{ needs.determine-tag.outputs.tag_name || 'latest' }}
run: |
vulkaninfo --summary
GG_BUILD_VULKAN=1 bash ./ci/run.sh ~/results/llama.cpp ~/mnt/llama.cpp

ggml-ci-linux-intel-vulkan:
needs: determine-tag
runs-on: [self-hosted, Linux, Intel]

steps:
Expand All @@ -237,11 +270,14 @@ jobs:

- name: Test
id: ggml-ci
env:
HF_WEBUI_VERSION: ${{ needs.determine-tag.outputs.tag_name || 'latest' }}
run: |
vulkaninfo --summary
GG_BUILD_VULKAN=1 bash ./ci/run.sh ~/results/llama.cpp ~/mnt/llama.cpp

ggml-ci-win-intel-vulkan:
needs: determine-tag
runs-on: [self-hosted, Windows, X64, Intel]

steps:
Expand All @@ -256,13 +292,15 @@ jobs:
MSYSTEM: UCRT64
CHERE_INVOKING: 1
PATH: C:\msys64\ucrt64\bin;C:\msys64\usr\bin;C:\Windows\System32;${{ env.PATH }}
HF_WEBUI_VERSION: ${{ needs.determine-tag.outputs.tag_name || 'latest' }}
run: |
vulkaninfo --summary
# Skip python related tests with GG_BUILD_LOW_PERF=1 since Windows MSYS2 UCRT64 currently fails to create
# a valid python environment for testing
LLAMA_FATAL_WARNINGS=OFF GG_BUILD_NINJA=1 GG_BUILD_VULKAN=1 GG_BUILD_LOW_PERF=1 ./ci/run.sh ./results/llama.cpp ./mnt/llama.cpp

ggml-ci-intel-openvino-gpu-low-perf:
needs: determine-tag
runs-on: [self-hosted, Linux, Intel, OpenVINO]

concurrency:
Expand Down Expand Up @@ -294,6 +332,8 @@ jobs:

- name: Test
id: ggml-ci
env:
HF_WEBUI_VERSION: ${{ needs.determine-tag.outputs.tag_name || 'latest' }}
run: |
source ./openvino_toolkit/setupvars.sh
GG_BUILD_OPENVINO=1 GGML_OPENVINO_DEVICE=GPU GG_BUILD_LOW_PERF=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
8 changes: 6 additions & 2 deletions tools/server/CMakeLists.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -76,12 +76,16 @@ if (LLAMA_BUILD_WEBUI)

# Priority 2: Build-time asset provisioning (npm build → HF Bucket fallback)
if(NOT WEBUI_SOURCE_DIR)
if(DEFINED LLAMA_BUILD_NUMBER)
# Environment variable takes precedence (e.g., from CI workflows)
if(DEFINED ENV{HF_WEBUI_VERSION})
set(HF_WEBUI_VERSION "$ENV{HF_WEBUI_VERSION}")
Comment thread
allozaur marked this conversation as resolved.
message(STATUS "WebUI: using HF_WEBUI_VERSION from environment=${HF_WEBUI_VERSION}")
elseif(DEFINED LLAMA_BUILD_NUMBER)
set(HF_WEBUI_VERSION "${LLAMA_BUILD_NUMBER}")
Comment thread
allozaur marked this conversation as resolved.
Outdated
message(STATUS "WebUI: using LLAMA_BUILD_NUMBER=${HF_WEBUI_VERSION}")
else()
set(HF_WEBUI_VERSION "")
message(STATUS "WebUI: LLAMA_BUILD_NUMBER not defined")
message(STATUS "WebUI: version not specified (will use HF 'latest')")
endif()

# Stamp file embeds the version tag so a changed build number triggers
Expand Down
Loading