mirror of
https://github.com/BillyOutlast/flash-attention-prebuild-wheels-rocm.git
synced 2026-07-01 01:27:54 -04:00
Junya Morioka
197f0e7693
Remove working-directory from Upload Release Asset step to ensure gh CLI uses the correct repository (flash-attention-prebuild-wheels) instead of the cloned flash-attention repository.
210 lines
8.0 KiB
YAML
210 lines
8.0 KiB
YAML
# #########################################################
|
|
# Build wheels with GitHub hosted runner on Windows x86_64
|
|
# #########################################################
|
|
# Prerequisites (must be pre-installed on the runner):
|
|
# - pwsh
|
|
# - Git
|
|
# - GitHub CLI (gh)
|
|
# - Chocolatey
|
|
# - Visual Studio BuildTools 2022 with:
|
|
# - Microsoft.VisualStudio.Component.VC.Tools.x86.x64
|
|
# - Microsoft.VisualStudio.Component.VC.CMake.Project
|
|
# - Microsoft.VisualStudio.Component.Windows11SDK.22621
|
|
# - vswhere
|
|
# - CMake
|
|
# - Ninja
|
|
# - Make (optional)
|
|
# #########################################################
|
|
|
|
name: "[Windows x86_64, self-hosted] Build wheels and upload to GitHub Releases"
|
|
|
|
on:
|
|
workflow_call:
|
|
inputs:
|
|
flash-attn-version:
|
|
description: "Flash-Attention version"
|
|
required: true
|
|
type: string
|
|
python-version:
|
|
description: "Python version"
|
|
required: true
|
|
type: string
|
|
torch-version:
|
|
description: "PyTorch version"
|
|
required: true
|
|
type: string
|
|
cuda-version:
|
|
description: "CUDA version"
|
|
required: true
|
|
type: string
|
|
runner:
|
|
description: "Runner type"
|
|
required: false
|
|
type: string
|
|
default: '["self-hosted", "windows", "x64"]'
|
|
is-upload:
|
|
description: "Whether to upload the release asset"
|
|
required: false
|
|
type: boolean
|
|
default: true
|
|
|
|
jobs:
|
|
build_windows_wheels_self_hosted:
|
|
name: Build wheels and Upload (Windows x86_64, self-hosted runner)
|
|
runs-on: ${{ fromjson(inputs.runner) }}
|
|
timeout-minutes: 2160
|
|
env:
|
|
MAX_JOBS: 2
|
|
NVCC_THREADS: 2
|
|
steps:
|
|
- uses: actions/checkout@v4
|
|
|
|
- name: Enable Git long paths
|
|
shell: pwsh
|
|
run: git config --system core.longpaths true
|
|
|
|
# Install Python using uv because setup-python needs newer version of glibc
|
|
- name: Install uv
|
|
uses: astral-sh/setup-uv@v7
|
|
|
|
- name: Install Python
|
|
shell: pwsh
|
|
run: |
|
|
uv venv -p ${{ inputs.python-version }}
|
|
uv pip install -U pip setuptools==75.8.0 wheel packaging psutil numpy ninja
|
|
$current_dir = (Get-Location).Path
|
|
echo "$current_dir\.venv\Scripts" >> $env:GITHUB_PATH
|
|
|
|
- uses: mjun0812/setup-cuda@v1
|
|
with:
|
|
version: ${{ inputs.cuda-version }}
|
|
|
|
# Visual Studio BuildTools is pre-installed on the runner
|
|
- name: Setup MSVC Developer Command Prompt
|
|
uses: TheMrMilchmann/setup-msvc-dev@v3
|
|
with:
|
|
arch: x64
|
|
|
|
- name: Add msbuild to PATH
|
|
uses: microsoft/setup-msbuild@v2
|
|
|
|
- name: Build wheels
|
|
shell: pwsh
|
|
timeout-minutes: 2160
|
|
run: |
|
|
.\build_windows.ps1 -FlashAttnVersion "${{ inputs.flash-attn-version }}" -PythonVersion "${{ inputs.python-version }}" -TorchVersion "${{ inputs.torch-version }}" -CudaVersion "${{ inputs.cuda-version }}"
|
|
# Script changes to flash-attention directory, wheel is in dist/
|
|
$wheelName = Get-ChildItem -Path "dist\*.whl" | Select-Object -First 1 | ForEach-Object { $_.Name }
|
|
echo "wheel_name=$wheelName" >> $env:GITHUB_ENV
|
|
|
|
- name: Install Test
|
|
shell: pwsh
|
|
working-directory: flash-attention
|
|
run: |
|
|
pip install --no-cache-dir dist/$env:wheel_name
|
|
python -c "import flash_attn; print(flash_attn.__version__)"
|
|
|
|
- name: Upload Release Asset
|
|
if: ${{ inputs.is-upload }}
|
|
shell: pwsh
|
|
env:
|
|
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
|
|
run: |
|
|
$tag_name = "${{ github.ref_name }}"
|
|
$wheel_path = "flash-attention/dist/$env:wheel_name"
|
|
gh release upload "$tag_name" "$wheel_path" --clobber
|
|
|
|
# Cleanup step - always runs even if previous steps fail
|
|
# Only cleans up Python and CUDA installations (VS BuildTools is pre-installed)
|
|
- name: Cleanup (always run)
|
|
if: always()
|
|
shell: pwsh
|
|
run: |
|
|
Write-Host "=========================================="
|
|
Write-Host "Starting cleanup for self-hosted runner..."
|
|
Write-Host "=========================================="
|
|
|
|
# 1. Remove flash-attention directory (source and build artifacts)
|
|
$flashAttnDir = Join-Path (Get-Location) "flash-attention"
|
|
if (Test-Path $flashAttnDir) {
|
|
Write-Host "[1/6] Removing flash-attention directory: $flashAttnDir"
|
|
Remove-Item -Path $flashAttnDir -Recurse -Force -ErrorAction SilentlyContinue
|
|
} else {
|
|
Write-Host "[1/6] flash-attention directory not found, skipping"
|
|
}
|
|
|
|
# 2. Remove Python virtual environment (.venv)
|
|
$venvDir = Join-Path (Get-Location) ".venv"
|
|
if (Test-Path $venvDir) {
|
|
Write-Host "[2/6] Removing Python virtual environment: $venvDir"
|
|
Remove-Item -Path $venvDir -Recurse -Force -ErrorAction SilentlyContinue
|
|
} else {
|
|
Write-Host "[2/6] .venv directory not found, skipping"
|
|
}
|
|
|
|
# 3. Remove pip cache
|
|
$pipCacheDir = Join-Path $env:LOCALAPPDATA "pip\cache"
|
|
if (Test-Path $pipCacheDir) {
|
|
Write-Host "[3/6] Removing pip cache: $pipCacheDir"
|
|
Remove-Item -Path $pipCacheDir -Recurse -Force -ErrorAction SilentlyContinue
|
|
} else {
|
|
Write-Host "[3/6] pip cache not found, skipping"
|
|
}
|
|
|
|
# 4. Remove uv cache
|
|
$uvCacheDir = Join-Path $env:LOCALAPPDATA "uv"
|
|
if (Test-Path $uvCacheDir) {
|
|
Write-Host "[4/6] Removing uv cache: $uvCacheDir"
|
|
Remove-Item -Path $uvCacheDir -Recurse -Force -ErrorAction SilentlyContinue
|
|
} else {
|
|
Write-Host "[4/6] uv cache not found, skipping"
|
|
}
|
|
|
|
# 5. Uninstall CUDA using proper Windows uninstaller
|
|
Write-Host "[5/6] Uninstalling CUDA packages..."
|
|
$cudaPackages = Get-Package -Name "*CUDA*" -ErrorAction SilentlyContinue
|
|
if ($cudaPackages) {
|
|
foreach ($package in $cudaPackages) {
|
|
Write-Host " Uninstalling: $($package.Name)"
|
|
try {
|
|
$package | Uninstall-Package -Force -ErrorAction SilentlyContinue
|
|
} catch {
|
|
Write-Host " Warning: Could not uninstall via Package Manager, trying alternative method..."
|
|
}
|
|
}
|
|
} else {
|
|
Write-Host " No CUDA packages found in Package Manager"
|
|
}
|
|
|
|
# Also try NVIDIA uninstaller if exists
|
|
$cudaBaseDir = "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA"
|
|
if (Test-Path $cudaBaseDir) {
|
|
Get-ChildItem -Path $cudaBaseDir -Directory -ErrorAction SilentlyContinue | ForEach-Object {
|
|
$uninstaller = Join-Path $_.FullName "Uninstall.exe"
|
|
if (Test-Path $uninstaller) {
|
|
Write-Host " Running CUDA uninstaller: $uninstaller"
|
|
Start-Process -FilePath $uninstaller -ArgumentList "/S" -Wait -ErrorAction SilentlyContinue
|
|
}
|
|
}
|
|
# Remove remaining directory if exists
|
|
if (Test-Path $cudaBaseDir) {
|
|
Write-Host " Removing remaining CUDA directory: $cudaBaseDir"
|
|
Remove-Item -Path $cudaBaseDir -Recurse -Force -ErrorAction SilentlyContinue
|
|
}
|
|
}
|
|
|
|
# 6. Remove temp files
|
|
Write-Host "[6/6] Removing temporary files"
|
|
$tempPatterns = @("pip-*", "torch*", "cuda*", "flash*", "uv-*")
|
|
foreach ($pattern in $tempPatterns) {
|
|
$tempPath = Join-Path $env:TEMP $pattern
|
|
Get-ChildItem -Path $tempPath -ErrorAction SilentlyContinue | ForEach-Object {
|
|
Write-Host " Removing temp: $($_.FullName)"
|
|
Remove-Item -Path $_.FullName -Recurse -Force -ErrorAction SilentlyContinue
|
|
}
|
|
}
|
|
|
|
Write-Host "=========================================="
|
|
Write-Host "Cleanup completed."
|
|
Write-Host "=========================================="
|