Skip to content
Linux_CI

Adding cuda kernel (optimized for sm80) for block-wise 4b quantized float 16 GEMM. #24407

Sign in to view logs

Adding cuda kernel (optimized for sm80) for block-wise 4b quantized float 16 GEMM.

Adding cuda kernel (optimized for sm80) for block-wise 4b quantized float 16 GEMM. #24407

Workflow file for this run

.github/workflows/linux.yml at cf39757
name: Linux_CI
on:
push:
branches:
- main
- rel-*
pull_request:
concurrency:
group: ${{ github.workflow }}-${{ github.head_ref || github.run_id }}
cancel-in-progress: true
jobs:
Onnxruntime-TVM:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
with:
submodules: true
- uses: actions/setup-python@v5
with:
python-version: '3.8.x'
architecture: 'x64'
- name: 'Setup TVM EP requirements'
run: |
set -e -x
sudo apt-get update
sudo apt-get install -y libtinfo-dev zlib1g-dev build-essential libedit-dev libxml2-dev nasm
python3 -m pip install -r ${{ github.workspace }}/tools/ci_build/github/linux/tvm/requirements.txt
- name: 'Build and Test'
run: |
python3 ${{ github.workspace }}/tools/ci_build/build.py --build_dir build --config Release --skip_submodule_sync --parallel --enable_pybind --disable_contrib_ops --disable_ml_ops --skip_onnx_tests --use_tvm --use_tvm_hash --ctest_path ""