Skip to content

[WebGPU EP] extend concat to handle large number of inputs #3116

[WebGPU EP] extend concat to handle large number of inputs

[WebGPU EP] extend concat to handle large number of inputs #3116

Workflow file for this run

# This workflow builds and tests the ONNX Runtime for Linux for DNNL EP
# It leverages a reusable workflow (`reusable_linux_build.yml`) to handle the core build and test logic
# within Docker containers, ensuring a consistent environment.
# This file is very similar to linux_ci.yml, but much simpler
name: Linux DNNL CI
on:
push:
branches: [main, 'rel-*']
pull_request:
branches: [main, 'rel-*']
workflow_dispatch:
concurrency:
group: ${{ github.workflow }}-${{ github.event_name == 'pull_request' && github.ref || github.sha }}
cancel-in-progress: true
permissions:
contents: read
packages: write
attestations: write
id-token: write
jobs:
build-linux-x64-release-dnnl:
name: Build Linux x64 Release (DNNL EP)
uses: ./.github/workflows/reusable_linux_build.yml
with:
pool_name: "onnxruntime-github-Ubuntu2204-AMD-CPU"
build_config: Release
architecture: x64
dockerfile_path: tools/ci_build/github/linux/docker/Dockerfile.manylinux2_28_cpu
docker_image_repo: onnxruntimecpubuildpythonx64
extra_build_flags: '--use_binskim_compliant_compile_flags --build_wheel --build_nuget --use_dnnl'
python_path_prefix: 'PATH=/opt/python/cp310-cp310/bin:$PATH'
secrets:
GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}