Files
LocalAI/backend/python
Isabel Wu 977f663cb0 fix(trl): disable inline GRPO reward code by default (RCE, #11015) (#11068)
fix(trl): disable inline GRPO reward code by default (RCE)

POST /api/fine-tuning/jobs accepts reward_functions[].code, an inline Python
body, and compile_inline_reward() execs it against a restricted-builtins
allowlist (_SAFE_BUILTINS). That allowlist is not a security boundary:
().__class__.__bases__[0].__subclasses__() reaches os._wrap_close and thus
os.system, giving arbitrary code execution. The fine-tuning endpoint is
unauthenticated by default, so any caller could run code on the host.

Hardening the allowlist is a losing game against CPython introspection, so
inline reward code is now refused unless the operator explicitly opts in with
LOCALAI_TRL_ALLOW_INLINE_REWARD=true on the backend. Builtin reward functions
are unaffected. The gate lives in build_reward_functions(), the single point
all inline specs flow through. Docs updated to stop describing the allowlist as
a sandbox and to document the opt-in.

Fixes #11015

Signed-off-by: Isabel Wu <231155141+wuisabel-gif@users.noreply.github.com>
Co-authored-by: Isabel Wu <231155141+wuisabel-gif@users.noreply.github.com>
2026-07-23 23:10:34 +02:00
..
2026-04-12 08:51:30 +02:00
2026-04-12 08:51:30 +02:00

Python Backends for LocalAI

This directory contains Python-based AI backends for LocalAI, providing support for various AI models and hardware acceleration targets.

Overview

The Python backends use a unified build system based on libbackend.sh that provides:

  • Automatic virtual environment management with support for both uv and pip
  • Hardware-specific dependency installation (CPU, CUDA, Intel, MLX, etc.)
  • Portable Python support for standalone deployments
  • Consistent backend execution across different environments

Available Backends

Core AI Models

  • transformers - Hugging Face Transformers framework (PyTorch-based)
  • vllm - High-performance LLM inference engine
  • mlx - Apple Silicon optimized ML framework

Audio & Speech

  • coqui - Coqui TTS models
  • faster-whisper - Fast Whisper speech recognition
  • kitten-tts - Lightweight TTS
  • mlx-audio - Apple Silicon audio processing
  • chatterbox - TTS model
  • kokoro - TTS models

Computer Vision

  • diffusers - Stable Diffusion and image generation
  • longcat-video - CUDA video and speech-driven avatar generation with LongCat-Video
  • mlx-vlm - Vision-language models for Apple Silicon
  • rfdetr - Object detection models

Specialized

  • rerankers - Text reranking models

Quick Start

Prerequisites

  • Python 3.10+ (default: 3.10.18)
  • uv package manager (recommended) or pip
  • Appropriate hardware drivers for your target (CUDA, Intel, etc.)

Installation

Each backend can be installed individually:

# Navigate to a specific backend
cd backend/python/transformers

# Install dependencies
make transformers
# or
bash install.sh

# Run the backend
make run
# or
bash run.sh

Using the Unified Build System

The libbackend.sh script provides consistent commands across all backends:

# Source the library in your backend script
source $(dirname $0)/../common/libbackend.sh

# Install requirements (automatically handles hardware detection)
installRequirements

# Start the backend server
startBackend $@

# Run tests
runUnittests

Hardware Targets

The build system automatically detects and configures for different hardware:

  • CPU - Standard CPU-only builds
  • CUDA - NVIDIA GPU acceleration (supports CUDA 12/13)
  • Intel - Intel XPU/GPU optimization
  • MLX - Apple Silicon (M1/M2/M3) optimization
  • HIP - AMD GPU acceleration

Target-Specific Requirements

Backends can specify hardware-specific dependencies:

  • requirements.txt - Base requirements
  • requirements-cpu.txt - CPU-specific packages
  • requirements-cublas12.txt - CUDA 12 packages
  • requirements-cublas13.txt - CUDA 13 packages
  • requirements-intel.txt - Intel-optimized packages
  • requirements-mps.txt - Apple Silicon packages

Configuration Options

Environment Variables

  • PYTHON_VERSION - Python version (default: 3.10)
  • PYTHON_PATCH - Python patch version (default: 18)
  • BUILD_TYPE - Force specific build target
  • USE_PIP - Use pip instead of uv (default: false)
  • PORTABLE_PYTHON - Enable portable Python builds
  • LIMIT_TARGETS - Restrict backend to specific targets

Example: CUDA 12 Only Backend

# In your backend script
LIMIT_TARGETS="cublas12"
source $(dirname $0)/../common/libbackend.sh

Example: Intel-Optimized Backend

# In your backend script
LIMIT_TARGETS="intel"
source $(dirname $0)/../common/libbackend.sh

Development

Adding a New Backend

  1. Create a new directory in backend/python/
  2. Copy the template structure from common/template/
  3. Implement your backend.py with the required gRPC interface
  4. Add appropriate requirements files for your target hardware
  5. Use libbackend.sh for consistent build and execution

Testing

# Run backend tests
make test
# or
bash test.sh

Building

# Install dependencies
make <backend-name>

# Clean build artifacts
make clean

Architecture

Each backend follows a consistent structure:

backend-name/
├── backend.py          # Main backend implementation
├── requirements.txt    # Base dependencies
├── requirements-*.txt  # Hardware-specific dependencies
├── install.sh         # Installation script
├── run.sh            # Execution script
├── test.sh           # Test script
├── Makefile          # Build targets
└── test.py           # Unit tests

Troubleshooting

Common Issues

  1. Missing dependencies: Ensure all requirements files are properly configured
  2. Hardware detection: Check that BUILD_TYPE matches your system
  3. Python version: Verify Python 3.10+ is available
  4. Virtual environment: Use ensureVenv to create/activate environments

Contributing

When adding new backends or modifying existing ones:

  1. Follow the established directory structure
  2. Use libbackend.sh for consistent behavior
  3. Include appropriate requirements files for all target hardware
  4. Add comprehensive tests
  5. Update this README if adding new backend types