Installation¶
From source¶
Clone the repository and install in development mode:
git clone git@github.com:kevinkorfmann/cxt.git
cd cxt
pip install -e .
Checkpoints are downloaded automatically on first use via
cxt.load_model() and cached in ~/.cache/cxt/checkpoints/.
To redirect the cache (e.g. for isolated reproduction runs), set the
CXT_CHECKPOINT_CACHE environment variable:
export CXT_CHECKPOINT_CACHE=/path/to/my/checkpoints
If you prefer to have checkpoints available offline via Git LFS:
conda install anaconda::git-lfs # or: apt install git-lfs
git lfs install
git lfs pull
Requirements¶
Python 3.10+
PyTorch 2.0+
Lightning (for training)
msprime, tskit, stdpopsim (for simulation)
numpy, scipy, pandas, einops, tqdm
A CUDA-capable GPU is strongly recommended for inference. Multi-GPU setups (2–4 GPUs) provide near-linear speedups.
Optional dependencies¶
tszip– for compressed tree-sequence storage (human and mosquito examples)stdpopsim– for species-specific simulation scenariosmatplotlib– for plotting examplesrequests– for automatic checkpoint downloads
Verifying the installation¶
import cxt
model = cxt.load_model("broad", device="cpu")
print(f"Loaded model with {sum(p.numel() for p in model.parameters()):,} parameters")