Installation¶

Get ChimeraLM installed on your system in just a few minutes.

Prerequisites¶

Requirements

Python: 3.10 or 3.11 (3.12 not yet supported)
Operating System: Linux, macOS, or Windows
Optional: CUDA-compatible GPU for accelerated inference

Installation Methods¶

Choose your preferred installation method:

pipcondauvfrom source

The easiest way to install ChimeraLM:

pip install chimeralm

Verify the installation:

chimeralm --version

Install using conda:

conda install -c conda-forge chimeralm

Or create a new environment:

conda create -n chimeralm python=3.10
conda activate chimeralm
pip install chimeralm

Using the fast uv package manager:

uv pip install chimeralm

Or with a virtual environment:

uv venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate
uv pip install chimeralm

For development or the latest features:

# Clone the repository
git clone https://github.com/ylab-hi/ChimeraLM.git
cd ChimeraLM

# Install in development mode with uv
uv sync

# Verify installation
uv run chimeralm --version

GPU Setup¶

GPU Acceleration

For significantly faster predictions, install GPU support:

CUDA (NVIDIA GPUs)¶

ChimeraLM automatically uses CUDA if available. Verify your GPU is detected:

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

If CUDA is not detected, install PyTorch with CUDA support:

pip install torch==2.5.1 --index-url https://download.pytorch.org/whl/cu121

MPS (Apple Silicon)¶

For M1/M2/M3 Macs, MPS acceleration is automatically enabled:

python -c "import torch; print(f'MPS available: {torch.backends.mps.is_available()}')"

Python Version Compatibility

ChimeraLM requires Python 3.10, 3.11 and 3.12. Python 3.13 is not yet supported due to dependency constraints.

Verification¶

Confirm ChimeraLM is installed correctly:

# Check version
chimeralm --version

# View available commands
chimeralm --help

# Test import (Python)
python -c "import chimeralm; print('ChimeraLM imported successfully')"

Expected output:

ChimeraLM imported successfully

Troubleshooting Installation¶

Common Issues¶

ImportError: No module named 'chimeralm'

Solution: Ensure you've activated the correct Python environment:

# Check which Python is being used
which python

# Reinstall in the current environment
pip install --force-reinstall chimeralm

CUDA out of memory during installation

Solution: This is normal if testing GPU during install. Reduce batch size when running predictions.

PyTorch version conflict

Solution: ChimeraLM requires PyTorch 2.5.1. Uninstall conflicting versions:

pip uninstall torch torchvision torchaudio
pip install chimeralm

Permission denied errors

Solution: Install in user space without sudo:

pip install --user chimeralm

For more issues, see the Troubleshooting Guide.

Next Steps¶

Now that ChimeraLM is installed, try the Quick Start tutorial to run your first prediction!