Build vLLM nightly wheels for CUDA 13.0 (#163239)

Now that https://github.com/vllm-project/vllm/pull/24599 has been merged
Pull Request resolved: https://github.com/pytorch/pytorch/pull/163239
Approved by: https://github.com/malfet, https://github.com/atalman
This commit is contained in:
Huy Do
2025-10-16 01:03:26 +00:00
committed by PyTorch MergeBot
parent ca8bd5dbed
commit c2bd41ac9f
2 changed files with 17 additions and 4 deletions

View File

@ -65,7 +65,7 @@ runs:
cd .ci/lumen_cli cd .ci/lumen_cli
python3 -m pip install -e . python3 -m pip install -e .
) )
MAX_JOBS="$(nproc --ignore=6)" MAX_JOBS="$(nproc --ignore=10)"
export MAX_JOBS export MAX_JOBS
# Split the comma-separated list and build each target # Split the comma-separated list and build each target

View File

@ -27,9 +27,8 @@ jobs:
fail-fast: false fail-fast: false
matrix: matrix:
python-version: [ '3.12' ] python-version: [ '3.12' ]
# TODO (huydhn): Add cu130 after https://github.com/vllm-project/vllm/issues/24464 is resolved
platform: [ 'manylinux_2_28_x86_64', 'manylinux_2_28_aarch64' ] platform: [ 'manylinux_2_28_x86_64', 'manylinux_2_28_aarch64' ]
device: [ 'cu128', 'cu129' ] device: [ 'cu128', 'cu129', 'cu130' ]
include: include:
- platform: manylinux_2_28_x86_64 - platform: manylinux_2_28_x86_64
device: cu128 device: cu128
@ -39,6 +38,10 @@ jobs:
device: cu129 device: cu129
manylinux-image: 'pytorch/manylinux2_28-builder:cuda12.9' manylinux-image: 'pytorch/manylinux2_28-builder:cuda12.9'
runner: linux.12xlarge.memory runner: linux.12xlarge.memory
- platform: manylinux_2_28_x86_64
device: cu130
manylinux-image: 'pytorch/manylinux2_28-builder:cuda13.0'
runner: linux.12xlarge.memory
- platform: manylinux_2_28_aarch64 - platform: manylinux_2_28_aarch64
device: cu128 device: cu128
manylinux-image: 'pytorch/manylinuxaarch64-builder:cuda12.8' manylinux-image: 'pytorch/manylinuxaarch64-builder:cuda12.8'
@ -47,6 +50,11 @@ jobs:
device: cu129 device: cu129
manylinux-image: 'pytorch/manylinuxaarch64-builder:cuda12.9' manylinux-image: 'pytorch/manylinuxaarch64-builder:cuda12.9'
runner: linux.arm64.r7g.12xlarge.memory runner: linux.arm64.r7g.12xlarge.memory
exclude:
# TODO (huydhn): Add cu130 aarch64 once PyTorch is on 2.9+ and
# xformers is update to support 13.0
- platform: manylinux_2_28_aarch64
device: cu130
name: "Build ${{ matrix.device }} vLLM wheel on ${{ matrix.platform }}" name: "Build ${{ matrix.device }} vLLM wheel on ${{ matrix.platform }}"
runs-on: ${{ matrix.runner }} runs-on: ${{ matrix.runner }}
timeout-minutes: 480 timeout-minutes: 480
@ -169,7 +177,12 @@ jobs:
fail-fast: false fail-fast: false
matrix: matrix:
platform: [ 'manylinux_2_28_x86_64', 'manylinux_2_28_aarch64' ] platform: [ 'manylinux_2_28_x86_64', 'manylinux_2_28_aarch64' ]
device: [ 'cu128', 'cu129' ] device: [ 'cu128', 'cu129', 'cu130' ]
exclude:
# TODO (huydhn): Add cu130 aarch64 once PyTorch is on 2.9+ and
# xformers is update to support 13.0
- platform: manylinux_2_28_aarch64
device: cu130
env: env:
PLATFORM: ${{ matrix.platform }} PLATFORM: ${{ matrix.platform }}
BUILD_DEVICE: ${{ matrix.device }} BUILD_DEVICE: ${{ matrix.device }}