streamer-tools/local-transcription
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Simplify build process: CUDA support now included by default

Browse Source 
Since pyproject.toml is configured to use PyTorch CUDA index by default,
all builds automatically include CUDA support. Removed redundant separate
CUDA build scripts and updated documentation.

Changes:
- Removed build-cuda.sh and build-cuda.bat (no longer needed)
- Updated build.sh and build.bat to include CUDA by default
  - Added "uv sync" step to ensure CUDA PyTorch is installed
  - Updated messages to clarify CUDA support is included
- Updated BUILD.md to reflect simplified build process
  - Removed separate CUDA build sections
  - Clarified all builds include CUDA support
  - Updated GPU support section
- Updated CLAUDE.md with simplified build commands

Benefits:
- Simpler build process (one script per platform instead of two)
- Less confusion about which script to use
- All builds work on any system (GPU or CPU)
- Automatic fallback to CPU if no GPU available
- pyproject.toml is single source of truth for dependencies

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
This commit is contained in:
jknapp
2025-12-28 19:09:36 -08:00
parent be53f2e962
commit d34d272cf0
6 changed files with 42 additions and 186 deletions
Download Patch File Download Diff File Expand all files Collapse all files

70
BUILD.md
View File

@@ -10,7 +10,7 @@ This guide explains how to build standalone executables for Linux and Windows.
## Building for Linux
### Standard Build (CPU-only):
### Standard Build (includes CUDA support):
```bash
# Make the build script executable (first time only)
@@ -20,20 +20,8 @@ chmod +x build.sh
./build.sh
```
### CUDA Build (GPU Support):
Build with CUDA support even without NVIDIA hardware:
```bash
# Make the build script executable (first time only)
chmod +x build-cuda.sh
# Run the CUDA build script
./build-cuda.sh
```
This will:
- Install PyTorch with CUDA 12.1 support
- Install PyTorch with CUDA 12.1 support (configured in pyproject.toml)
- Bundle CUDA runtime libraries (~600MB extra)
- Create an executable that works on both GPU and CPU systems
- Automatically fall back to CPU if no CUDA GPU is available
@@ -45,6 +33,12 @@ The executable will be created in `dist/LocalTranscription/LocalTranscription`
# Clean previous builds
rm -rf build dist
# Sync dependencies (includes CUDA PyTorch)
uv sync
# Remove incompatible enum34 package
uv pip uninstall -q enum34
# Build with PyInstaller
uv run pyinstaller local-transcription.spec
```
@@ -57,24 +51,15 @@ tar -czf LocalTranscription-Linux.tar.gz LocalTranscription/
## Building for Windows
### Standard Build (CPU-only):
### Standard Build (includes CUDA support):
```cmd
# Run the build script
build.bat
```
### CUDA Build (GPU Support):
Build with CUDA support even without NVIDIA hardware:
```cmd
# Run the CUDA build script
build-cuda.bat
```
This will:
- Install PyTorch with CUDA 12.1 support
- Install PyTorch with CUDA 12.1 support (configured in pyproject.toml)
- Bundle CUDA runtime libraries (~600MB extra)
- Create an executable that works on both GPU and CPU systems
- Automatically fall back to CPU if no CUDA GPU is available
@@ -87,6 +72,12 @@ The executable will be created in `dist\LocalTranscription\LocalTranscription.ex
rmdir /s /q build
rmdir /s /q dist
# Sync dependencies (includes CUDA PyTorch)
uv sync
# Remove incompatible enum34 package
uv pip uninstall -q enum34
# Build with PyInstaller
uv run pyinstaller local-transcription.spec
```
@@ -129,7 +120,7 @@ By default, the console window is visible (for debugging). To hide it:
### GPU Support
#### Building with CUDA (Recommended for Distribution)
**CUDA support is included by default** in all builds via the PyTorch CUDA configuration in `pyproject.toml`.
**Yes, you CAN build with CUDA support on systems without NVIDIA GPUs!**
@@ -140,41 +131,16 @@ PyTorch provides CUDA-enabled builds that bundle the CUDA runtime libraries. Thi
3. **Automatic fallback** - the app detects available hardware and uses GPU if available, CPU otherwise
4. **Larger file size** - adds ~600MB-1GB to the executable size
**How it works:**
```bash
# Linux
./build-cuda.sh
# Windows
build-cuda.bat
```
The build script will:
- Install PyTorch with bundled CUDA 12.1 runtime
- Package all CUDA libraries into the executable
- Create a universal build that runs on any system
**When users run the executable:**
- If they have an NVIDIA GPU with drivers: Uses GPU acceleration
- If they don't have NVIDIA GPU: Automatically uses CPU
- No configuration needed - it just works!
#### Alternative: CPU-Only Builds
If you only want CPU support (smaller file size):
```bash
# Linux
./build.sh
# Windows
build.bat
```
#### AMD GPU Support
- **ROCm**: Requires special PyTorch builds from AMD
- Not recommended for general distribution
- Better to use CUDA build (works on all systems) or CPU build
- The default CUDA build already works on all systems (NVIDIA GPU, AMD GPU, or CPU-only)
### Optimizations

14
CLAUDE.md
View File

@@ -64,23 +64,19 @@ uv pip install torch --index-url https://download.pytorch.org/whl/cu121
### Building Executables
```bash
# Linux (CPU-only)
# Linux (includes CUDA support - works on both GPU and CPU systems)
./build.sh
# Linux (with CUDA support - works on both GPU and CPU systems)
./build-cuda.sh
# Windows (CPU-only)
# Windows (includes CUDA support - works on both GPU and CPU systems)
build.bat
# Windows (with CUDA support)
build-cuda.bat
# Manual build with PyInstaller
uv sync # Install dependencies (includes CUDA PyTorch)
uv pip uninstall -q enum34 # Remove incompatible enum34 package
uv run pyinstaller local-transcription.spec
```
**Important:** CUDA builds can be created on systems without NVIDIA GPUs. The PyTorch CUDA runtime is bundled, and the app automatically falls back to CPU if no GPU is available.
**Important:** All builds include CUDA support via `pyproject.toml` configuration. CUDA builds can be created on systems without NVIDIA GPUs. The PyTorch CUDA runtime is bundled, and the app automatically falls back to CPU if no GPU is available.
### Testing
```bash

61
build-cuda.bat
View File

@@ -1,61 +0,0 @@
@echo off
REM Build script for Windows with CUDA support
echo Building Local Transcription with CUDA support...
echo ==================================================
echo.
echo This will create a build that supports both CPU and CUDA GPUs.
echo The executable will be larger (~2-3GB) but will work on any system.
echo.
set /p INSTALL_CUDA="Install PyTorch with CUDA support? (y/n) "
if /i "%INSTALL_CUDA%"=="y" (
echo Installing PyTorch with CUDA 12.1 support...
REM Uninstall CPU-only version if present
REM Note: uv doesn't support -y flag, it uninstalls without confirmation
uv pip uninstall torch 2>nul
REM Install CUDA-enabled PyTorch
REM This installs PyTorch with bundled CUDA runtime
uv pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu121
echo CUDA-enabled PyTorch installed
echo.
)
REM Clean previous builds
echo Cleaning previous builds...
if exist build rmdir /s /q build
if exist dist rmdir /s /q dist
REM Remove enum34 if present (incompatible with PyInstaller)
echo Removing enum34 (if present)...
uv pip uninstall -q enum34 2>nul
REM Build with PyInstaller
echo Running PyInstaller...
uv run pyinstaller local-transcription.spec
REM Check if build succeeded
if exist "dist\LocalTranscription" (
echo.
echo Build successful!
echo Executable location: dist\LocalTranscription\LocalTranscription.exe
echo.
echo CUDA Support: YES (falls back to CPU if CUDA not available^)
echo.
echo To run the application:
echo cd dist\LocalTranscription
echo LocalTranscription.exe
echo.
echo To create a distributable package:
echo - Compress the dist\LocalTranscription folder to a ZIP file
echo - Name it: LocalTranscription-Windows-CUDA.zip
echo.
echo Note: This build will work on systems with or without NVIDIA GPUs.
) else (
echo.
echo Build failed!
exit /b 1
)

62
build-cuda.sh
View File

@@ -1,62 +0,0 @@
#!/bin/bash
# Build script for Linux with CUDA support
echo "Building Local Transcription with CUDA support..."
echo "=================================================="
echo ""
echo "This will create a build that supports both CPU and CUDA GPUs."
echo "The executable will be larger (~2-3GB) but will work on any system."
echo ""
# Check if we should install CUDA-enabled PyTorch
read -p "Install PyTorch with CUDA support? (y/n) " -n 1 -r
echo
if [[ $REPLY =~ ^[Yy]$ ]]
then
echo "Installing PyTorch with CUDA 12.1 support..."
# Uninstall CPU-only version if present
# Note: uv doesn't support -y flag, it uninstalls without confirmation
uv pip uninstall torch 2>/dev/null || true
# Install CUDA-enabled PyTorch
# This installs PyTorch with bundled CUDA runtime
uv pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu121
echo "✓ CUDA-enabled PyTorch installed"
echo ""
fi
# Clean previous builds
echo "Cleaning previous builds..."
rm -rf build dist
# Remove enum34 if present (incompatible with PyInstaller)
echo "Removing enum34 (if present)..."
uv pip uninstall -q enum34 2>/dev/null || true
# Build with PyInstaller
echo "Running PyInstaller..."
uv run pyinstaller local-transcription.spec
# Check if build succeeded
if [ -d "dist/LocalTranscription" ]; then
echo ""
echo "✓ Build successful!"
echo "Executable location: dist/LocalTranscription/LocalTranscription"
echo ""
echo "CUDA Support: YES (falls back to CPU if CUDA not available)"
echo ""
echo "To run the application:"
echo " cd dist/LocalTranscription"
echo " ./LocalTranscription"
echo ""
echo "To create a distributable package:"
echo " cd dist"
echo " tar -czf LocalTranscription-Linux-CUDA.tar.gz LocalTranscription/"
echo ""
echo "Note: This build will work on systems with or without NVIDIA GPUs."
else
echo ""
echo "✗ Build failed!"
exit 1
fi

10
build.bat
View File

@@ -1,15 +1,21 @@
@echo off
REM Build script for Windows
REM Build script for Windows with CUDA support (falls back to CPU if no GPU)
echo Building Local Transcription for Windows...
echo ==========================================
echo.
echo This build includes CUDA support and works on both GPU and CPU systems.
echo.
REM Clean previous builds
echo Cleaning previous builds...
if exist build rmdir /s /q build
if exist dist rmdir /s /q dist
REM Sync dependencies (uses PyTorch CUDA from pyproject.toml)
echo Installing dependencies with CUDA support...
uv sync
REM Remove enum34 if present (incompatible with PyInstaller)
echo Removing enum34 (if present)...
uv pip uninstall -q enum34 2>nul
@@ -24,6 +30,8 @@ if exist "dist\LocalTranscription" (
echo Build successful!
echo Executable location: dist\LocalTranscription\LocalTranscription.exe
echo.
echo CUDA Support: YES (automatically falls back to CPU if no GPU detected^)
echo.
echo To run the application:
echo cd dist\LocalTranscription
echo LocalTranscription.exe

11
build.sh
View File

@@ -1,13 +1,20 @@
#!/bin/bash
# Build script for Linux
# Build script for Linux with CUDA support (falls back to CPU if no GPU)
echo "Building Local Transcription for Linux..."
echo "========================================="
echo ""
echo "This build includes CUDA support and works on both GPU and CPU systems."
echo ""
# Clean previous builds
echo "Cleaning previous builds..."
rm -rf build dist
# Sync dependencies (uses PyTorch CUDA from pyproject.toml)
echo "Installing dependencies with CUDA support..."
uv sync
# Remove enum34 if present (incompatible with PyInstaller)
echo "Removing enum34 (if present)..."
uv pip uninstall -q enum34 2>/dev/null || true
@@ -22,6 +29,8 @@ if [ -d "dist/LocalTranscription" ]; then
echo "✓ Build successful!"
echo "Executable location: dist/LocalTranscription/LocalTranscription"
echo ""
echo "CUDA Support: YES (automatically falls back to CPU if no GPU detected)"
echo ""
echo "To run the application:"
echo " cd dist/LocalTranscription"
echo " ./LocalTranscription"