Installation Guide

This guide walks you through installing MAPLE and its dependencies on your system. MAPLE supports Linux and macOS, with optional GPU acceleration via CUDA.

Requirements

Before installing MAPLE, ensure your system meets the following minimum requirements:

Install MAPLE

Clone the MAPLE repository and install it in development mode:

# Clone the repository
git clone https://github.com/MAPLE-project/MAPLE.git
cd MAPLE

# Install in editable/development mode
pip install -e .

This installs MAPLE as a Python package and makes the maple command available in your terminal. The -e flag enables editable mode, so any changes to the source code take effect immediately without reinstalling.

Install Dependencies

Core Dependencies

The core scientific libraries are installed automatically with pip install -e ., but you can also install them manually:

pip install numpy scipy matplotlib ase

PyTorch

MAPLE requires PyTorch for ML potential inference. Install the version that matches your hardware:

With CUDA support (recommended for GPU):

pip install torch --index-url https://download.pytorch.org/whl/cu121

CPU only:

pip install torch --index-url https://download.pytorch.org/whl/cpu

ML Potentials

To use specific ML potential models (ANI, AIMNet2, MACE, etc.), install the fairchem-core package:

pip install fairchem-core

This provides access to the following models:

• ANI family: ani2x, ani1x, ani1ccx, ani1xnr
• AIMNet2: aimnet2
• MACE-OFF23: maceoff23s, maceoff23m, maceoff23l
• Other: egret, uma

Verify Installation

After installation, verify that MAPLE is correctly set up by running:

maple --help

You should see the MAPLE help output listing available commands and options. If the maple command is not found, ensure your Python environment's bin directory is on your PATH.

You can also verify GPU availability:

python -c "import torch; print('CUDA available:', torch.cuda.is_available())"