mirror of
https://github.com/zebrajr/pytorch.git
synced 2026-01-15 12:15:51 +00:00
5587e43ea2
This matches the current CI setup on vLLM on CUDA 12.9, avoid any funny business between 12.8 and 12.9 Pull Request resolved: https://github.com/pytorch/pytorch/pull/170513 Approved by: https://github.com/yangw-dev, https://github.com/zou3519
66 lines
3.4 KiB
YAML
66 lines
3.4 KiB
YAML
name: vllm-test
|
|
|
|
on:
|
|
push:
|
|
branches:
|
|
- main
|
|
- release/*
|
|
tags:
|
|
- ciflow/vllm/*
|
|
workflow_dispatch:
|
|
schedule:
|
|
- cron: '0 */8 * * *' # every 8 hours at minute 0 (UTC)
|
|
|
|
concurrency:
|
|
group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref_name }}-${{ github.ref_type == 'branch' && github.sha }}-${{ github.event_name == 'workflow_dispatch' }}
|
|
cancel-in-progress: true
|
|
|
|
permissions:
|
|
id-token: write
|
|
contents: read
|
|
|
|
jobs:
|
|
build:
|
|
name: vllm-x-pytorch-build
|
|
if: github.repository_owner == 'pytorch'
|
|
uses: ./.github/workflows/_linux-build.yml
|
|
with:
|
|
# When building vLLM, uv doesn't like that we rename wheel without changing the wheel metadata
|
|
allow-reuse-old-whl: false
|
|
build-additional-packages: "vision audio"
|
|
build-external-packages: "vllm"
|
|
build-environment: linux-jammy-cuda12.9-py3.12-gcc11
|
|
docker-image-name: ci-image:pytorch-linux-jammy-cuda12.9-cudnn9-py3.12-gcc11-vllm
|
|
cuda-arch-list: '8.0 8.9 9.0'
|
|
runner: linux.24xlarge.memory
|
|
test-matrix: |
|
|
{ include: [
|
|
{ config: "vllm_basic_correctness_test", shard: 1, num_shards: 1, runner: "linux.g6.4xlarge.experimental.nvidia.gpu" },
|
|
{ config: "vllm_basic_models_test", shard: 1, num_shards: 1, runner: "linux.g6.4xlarge.experimental.nvidia.gpu" },
|
|
{ config: "vllm_entrypoints_test", shard: 1, num_shards: 1,runner: "linux.g6.4xlarge.experimental.nvidia.gpu" },
|
|
{ config: "vllm_regression_test", shard: 1, num_shards: 1, runner: "linux.g6.4xlarge.experimental.nvidia.gpu" },
|
|
{ config: "vllm_multi_model_processor_test", shard: 1, num_shards: 1, runner: "linux.g6.4xlarge.experimental.nvidia.gpu" },
|
|
{ config: "vllm_pytorch_compilation_unit_tests", shard: 1, num_shards: 1, runner: "linux.g6.4xlarge.experimental.nvidia.gpu" },
|
|
{ config: "vllm_lora_28_failure_test", shard: 1, num_shards: 1, runner: "linux.g6.4xlarge.experimental.nvidia.gpu" },
|
|
{ config: "vllm_multi_model_test_28_failure_test", shard: 1, num_shards: 1, runner: "linux.g6.4xlarge.experimental.nvidia.gpu"},
|
|
{ config: "vllm_language_model_test_extended_generation_28_failure_test", shard: 1, num_shards: 1, runner: "linux.g6.4xlarge.experimental.nvidia.gpu"},
|
|
{ config: "vllm_distributed_test_2_gpu_28_failure_test", shard: 1, num_shards: 1, runner: "linux.g6.4xlarge.experimental.nvidia.gpu" },
|
|
{ config: "vllm_lora_test", shard: 0, num_shards: 4, runner: "linux.g6.4xlarge.experimental.nvidia.gpu" },
|
|
{ config: "vllm_lora_test", shard: 1, num_shards: 4, runner: "linux.g6.4xlarge.experimental.nvidia.gpu" },
|
|
{ config: "vllm_lora_test", shard: 2, num_shards: 4, runner: "linux.g6.4xlarge.experimental.nvidia.gpu" },
|
|
{ config: "vllm_lora_test", shard: 3, num_shards: 4, runner: "linux.g6.4xlarge.experimental.nvidia.gpu" },
|
|
{ config: "vllm_lora_tp_test_distributed", shard: 1, num_shards: 1, runner: "linux.g6.12xlarge.nvidia.gpu"},
|
|
{ config: "vllm_distributed_test_28_failure_test", shard: 1, num_shards: 1, runner: "linux.g6.12xlarge.nvidia.gpu"}
|
|
]}
|
|
secrets: inherit
|
|
|
|
test:
|
|
name: vllm-x-pytorch-test
|
|
uses: ./.github/workflows/_linux-test.yml
|
|
needs: build
|
|
with:
|
|
build-environment: linux-jammy-cuda12.9-py3.12-gcc11
|
|
docker-image: ${{ needs.build.outputs.docker-image }}
|
|
test-matrix: ${{ needs.build.outputs.test-matrix }}
|
|
secrets: inherit
|