Lumpiasty/llama.cpp
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

ci : Update OpenVINO to 2026.3, skip nemotron-h rollback test (#27292)

Browse Source 
* update to ov-2026.3, update device drivers

* ci: skip nemotron-h rollback test on OpenVINO

The OpenVINO backend does not support SSM_SCAN, so the Nemotron-H recurrent state rollback graph is split and cannot preserve the recurrent cache output shape. Keep the test enabled for other backends and retain the qwen35 OpenVINO rollback coverage.

---------

Co-authored-by: ravi9 <ravi.panchumarthy@intel.com>
This commit is contained in:
Zijun Yu
2026-08-18 12:02:22 +02:00
committed by GitHub
co-authored by ravi9
parent c0296022f3
commit 9d77fa1725
7 changed files with 35 additions and 35 deletions
Download Patch File Download Diff File Expand all files Collapse all files
.devops/openvino.Dockerfile
+10 -10
View File
@@ -1,18 +1,18 @@
ARG OPENVINO_VERSION_MAJOR=2026.2.1 ARG OPENVINO_VERSION_MAJOR=2026.3
ARG OPENVINO_VERSION_FULL=2026.2.1.21919.ede283a88e3 ARG OPENVINO_VERSION_FULL=2026.3.0.22451.bd8d6542e3c
ARG UBUNTU_VERSION=24.04 ARG UBUNTU_VERSION=24.04
# Intel GPU driver versions. https://github.com/intel/compute-runtime/releases # Intel GPU driver versions. https://github.com/intel/compute-runtime/releases
ARG IGC_VERSION=v2.36.3 ARG IGC_VERSION=v2.38.2
ARG IGC_VERSION_FULL=2_2.36.3+21719 ARG IGC_VERSION_FULL=2_2.38.2+22051
ARG COMPUTE_RUNTIME_VERSION=26.22.38646.4 ARG COMPUTE_RUNTIME_VERSION=26.27.39122.11
ARG COMPUTE_RUNTIME_VERSION_FULL=26.22.38646.4-0 ARG COMPUTE_RUNTIME_VERSION_FULL=26.27.39122.11-0
ARG IGDGMM_VERSION=22.10.0 ARG IGDGMM_VERSION=22.10.0
# Intel NPU driver versions. https://github.com/intel/linux-npu-driver/releases # Intel NPU driver versions. https://github.com/intel/linux-npu-driver/releases
ARG NPU_DRIVER_VERSION=v1.33.0 ARG NPU_DRIVER_VERSION=v1.35.0
ARG NPU_DRIVER_FULL=v1.33.0.20260529-26625960453 ARG NPU_DRIVER_FULL=v1.35.0.20260722-29947505341
ARG LIBZE1_VERSION=1.27.0-1~24.04~ppa2 ARG LIBZE1_VERSION=1.28.2-1~24.04~ppa1
# Optional proxy build arguments # Optional proxy build arguments
ARG http_proxy= ARG http_proxy=
@@ -170,7 +170,7 @@ RUN --mount=type=cache,target=/var/cache/intel-npu,sharing=locked \
fi; \ fi; \
DEB=/var/cache/intel-npu/libze1_${LIBZE1_VERSION}_amd64.deb; \ DEB=/var/cache/intel-npu/libze1_${LIBZE1_VERSION}_amd64.deb; \
if [ ! -f "$DEB" ]; then \ if [ ! -f "$DEB" ]; then \
wget -q -O "$DEB" https://snapshot.ppa.launchpadcontent.net/kobuk-team/intel-graphics/ubuntu/20260324T100000Z/pool/main/l/level-zero-loader/libze1_${LIBZE1_VERSION}_amd64.deb; \ wget -q -O "$DEB" https://snapshot.ppa.launchpadcontent.net/kobuk-team/intel-graphics/ubuntu/20260606T100000Z/pool/main/l/level-zero-loader/libze1_${LIBZE1_VERSION}_amd64.deb; \
fi; \ fi; \
mkdir /tmp/npu/ && cd /tmp/npu/ && tar -xf "$TGZ" && cp "$DEB" .; \ mkdir /tmp/npu/ && cd /tmp/npu/ && tar -xf "$TGZ" && cp "$DEB" .; \
apt-get update; \ apt-get update; \
.github/workflows/build-cache.yml
+5 -5
View File
@@ -40,9 +40,9 @@ jobs:
runs-on: ubuntu-24.04 runs-on: ubuntu-24.04
env: env:
# Sync versions in build.yml, build-self-hosted.yml, release.yml, build-cache.yml, .devops/openvino.Dockerfile # Sync versions in build-openvino.yml, build-self-hosted.yml, release.yml, build-cache.yml, .devops/openvino.Dockerfile
OPENVINO_VERSION_MAJOR: "2026.2.1" OPENVINO_VERSION_MAJOR: "2026.3"
OPENVINO_VERSION_FULL: "2026.2.1.21919.ede283a88e3" OPENVINO_VERSION_FULL: "2026.3.0.22451.bd8d6542e3c"
steps: steps:
- name: Clone - name: Clone
@@ -69,8 +69,8 @@ jobs:
env: env:
# Sync versions in build.yml, build-self-hosted.yml, release.yml, build-cache.yml, .devops/openvino.Dockerfile # Sync versions in build.yml, build-self-hosted.yml, release.yml, build-cache.yml, .devops/openvino.Dockerfile
OPENVINO_VERSION_MAJOR: "2026.2.1" OPENVINO_VERSION_MAJOR: "2026.3"
OPENVINO_VERSION_FULL: "2026.2.1.21919.ede283a88e3" OPENVINO_VERSION_FULL: "2026.3.0.22451.bd8d6542e3c"
steps: steps:
- name: Clone - name: Clone
.github/workflows/build-openvino.yml
+7 -7
View File
@@ -39,8 +39,8 @@ jobs:
env: env:
# Sync versions in build-openvino.yml, build-self-hosted.yml, release.yml, build-cache.yml, .devops/openvino.Dockerfile # Sync versions in build-openvino.yml, build-self-hosted.yml, release.yml, build-cache.yml, .devops/openvino.Dockerfile
OPENVINO_VERSION_MAJOR: "2026.2.1" OPENVINO_VERSION_MAJOR: "2026.3"
OPENVINO_VERSION_FULL: "2026.2.1.21919.ede283a88e3" OPENVINO_VERSION_FULL: "2026.3.0.22451.bd8d6542e3c"
steps: steps:
- name: Clone - name: Clone
@@ -81,7 +81,7 @@ jobs:
# TODO: fix and re-enable the `test-llama-archs` test below # TODO: fix and re-enable the `test-llama-archs` test below
run: | run: |
cd ${{ github.workspace }} cd ${{ github.workspace }}
ctest --test-dir build/ReleaseOV -L main -E "test-llama-archs" --verbose --timeout 2000 ctest --test-dir build/ReleaseOV -L main -E "test-llama-archs|test-recurrent-state-rollback-nemotron-h" --verbose --timeout 2000
- name: Test (GPU) - name: Test (GPU)
id: cmake_test_gpu id: cmake_test_gpu
@@ -89,15 +89,15 @@ jobs:
run: | run: |
cd ${{ github.workspace }} cd ${{ github.workspace }}
export GGML_OPENVINO_DEVICE=GPU export GGML_OPENVINO_DEVICE=GPU
ctest --test-dir build/ReleaseOV -L main -E "test-llama-archs" --verbose --timeout 3000 ctest --test-dir build/ReleaseOV -L main -E "test-llama-archs|test-recurrent-state-rollback-nemotron-h" --verbose --timeout 3000
openvino-windows-2022: openvino-windows-2022:
runs-on: windows-2022 runs-on: windows-2022
env: env:
# Sync versions in build-openvino.yml, build-self-hosted.yml, release.yml, build-cache.yml, .devops/openvino.Dockerfile # Sync versions in build-openvino.yml, build-self-hosted.yml, release.yml, build-cache.yml, .devops/openvino.Dockerfile
OPENVINO_VERSION_MAJOR: "2026.2.1" OPENVINO_VERSION_MAJOR: "2026.3"
OPENVINO_VERSION_FULL: "2026.2.1.21919.ede283a88e3" OPENVINO_VERSION_FULL: "2026.3.0.22451.bd8d6542e3c"
steps: steps:
- name: Clone - name: Clone
@@ -166,4 +166,4 @@ jobs:
call "%OPENVINO_ROOT%\setupvars.bat" call "%OPENVINO_ROOT%\setupvars.bat"
cd build cd build
ctest --test-dir ReleaseOV -L main -E "test-llama-archs" -C Release --verbose --timeout 3000 ctest --test-dir ReleaseOV -L main -E "test-llama-archs|test-recurrent-state-rollback-nemotron-h" -C Release --verbose --timeout 3000
.github/workflows/build-self-hosted.yml
+2 -2
View File
@@ -288,8 +288,8 @@ jobs:
env: env:
# Sync versions in build.yml, build-self-hosted.yml, release.yml, build-cache.yml, .devops/openvino.Dockerfile # Sync versions in build.yml, build-self-hosted.yml, release.yml, build-cache.yml, .devops/openvino.Dockerfile
OPENVINO_VERSION_MAJOR: "2026.2.1" OPENVINO_VERSION_MAJOR: "2026.3"
OPENVINO_VERSION_FULL: "2026.2.1.21919.ede283a88e3" OPENVINO_VERSION_FULL: "2026.3.0.22451.bd8d6542e3c"
steps: steps:
- name: Clone - name: Clone
.github/workflows/release.yml
+4 -4
View File
@@ -446,8 +446,8 @@ jobs:
env: env:
# Sync versions in build-openvino.yml, build-self-hosted.yml, release.yml, build-cache.yml, .devops/openvino.Dockerfile # Sync versions in build-openvino.yml, build-self-hosted.yml, release.yml, build-cache.yml, .devops/openvino.Dockerfile
OPENVINO_VERSION_MAJOR: "2026.2.1" OPENVINO_VERSION_MAJOR: "2026.3"
OPENVINO_VERSION_FULL: "2026.2.1.21919.ede283a88e3" OPENVINO_VERSION_FULL: "2026.3.0.22451.bd8d6542e3c"
steps: steps:
- name: Set OpenVINO version output - name: Set OpenVINO version output
@@ -562,8 +562,8 @@ jobs:
env: env:
# Sync versions in build-openvino.yml, build-self-hosted.yml, release.yml, build-cache.yml, .devops/openvino.Dockerfile # Sync versions in build-openvino.yml, build-self-hosted.yml, release.yml, build-cache.yml, .devops/openvino.Dockerfile
OPENVINO_VERSION_MAJOR: "2026.2.1" OPENVINO_VERSION_MAJOR: "2026.3"
OPENVINO_VERSION_FULL: "2026.2.1.21919.ede283a88e3" OPENVINO_VERSION_FULL: "2026.3.0.22451.bd8d6542e3c"
steps: steps:
- name: Set OpenVINO version output - name: Set OpenVINO version output
ci/run.sh
+1 -1
View File
@@ -190,7 +190,7 @@ if [ ! -z ${GG_BUILD_OPENVINO} ]; then
CMAKE_EXTRA="${CMAKE_EXTRA} -DGGML_OPENVINO=ON" CMAKE_EXTRA="${CMAKE_EXTRA} -DGGML_OPENVINO=ON"
# TODO: fix and re-enable the `test-llama-archs` test below # TODO: fix and re-enable the `test-llama-archs` test below
CTEST_EXTRA="-E test-llama-archs" CTEST_EXTRA="-E test-llama-archs|test-recurrent-state-rollback-nemotron-h"
fi fi
## helpers ## helpers
docs/backend/OPENVINO.md
+6 -6
View File
@@ -237,8 +237,8 @@ chmod +x ubuntu-llamacpp-ov-install.sh
# ============================================ # ============================================
set -euo pipefail set -euo pipefail
OPENVINO_VERSION_MAJOR="2026.2.1" OPENVINO_VERSION_MAJOR="2026.3"
OPENVINO_VERSION_FULL="2026.2.1.21919.ede283a88e3" OPENVINO_VERSION_FULL="2026.3.0.22451.bd8d6542e3c"
SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
OPENVINO_INSTALL_DIR="/opt/intel/openvino_${OPENVINO_VERSION_MAJOR}" OPENVINO_INSTALL_DIR="/opt/intel/openvino_${OPENVINO_VERSION_MAJOR}"
@@ -334,7 +334,7 @@ echo " ./build/ReleaseOV/bin/llama-cli -m model.gguf"
``` ```
> [!NOTE] > [!NOTE]
> The script pins OpenVINO `2026.2.1` via the `OPENVINO_VERSION_MAJOR` / `OPENVINO_VERSION_FULL` variables at the top — edit them to track a different release. > The script pins OpenVINO `2026.3` via the `OPENVINO_VERSION_MAJOR` / `OPENVINO_VERSION_FULL` variables at the top — edit them to track a different release.
</details> </details>
@@ -364,8 +364,8 @@ REM ============================================
REM llama.cpp OpenVINO Build Script (Ninja) REM llama.cpp OpenVINO Build Script (Ninja)
REM ============================================ REM ============================================
set "OPENVINO_VERSION_MAJOR=2026.2.1" set "OPENVINO_VERSION_MAJOR=2026.3"
set "OPENVINO_VERSION_FULL=2026.2.1.21919.ede283a88e3" set "OPENVINO_VERSION_FULL=2026.3.0.22451.bd8d6542e3c"
set "SCRIPT_DIR=%~dp0" set "SCRIPT_DIR=%~dp0"
set "VCPKG_DIR=C:\vcpkg" set "VCPKG_DIR=C:\vcpkg"
@@ -547,7 +547,7 @@ endlocal
``` ```
> [!NOTE] > [!NOTE]
> The script pins OpenVINO `2026.2.1` via the `OPENVINO_VERSION_MAJOR` / `OPENVINO_VERSION_FULL` variables at the top — edit them to track a different release. From any new shell, source the matching `setupvars` script via the junction — `call "C:\Intel\openvino\setupvars.bat"` from `cmd`, or `& "C:\Intel\openvino\setupvars.ps1"` from PowerShell. If `winget` cannot register Visual Studio Build Tools on first run, install them once manually and re-run the script from an elevated **Developer Command Prompt for VS 2022**. > The script pins OpenVINO `2026.3` via the `OPENVINO_VERSION_MAJOR` / `OPENVINO_VERSION_FULL` variables at the top — edit them to track a different release. From any new shell, source the matching `setupvars` script via the junction — `call "C:\Intel\openvino\setupvars.bat"` from `cmd`, or `& "C:\Intel\openvino\setupvars.ps1"` from PowerShell. If `winget` cannot register Visual Studio Build Tools on first run, install them once manually and re-run the script from an elevated **Developer Command Prompt for VS 2022**.
</details> </details>