# Requirements

For small datasets (~1000 molecules), it is possible to train models within a few minutes on a standard laptop with CPUs only. However, for larger datasets and larger Uni-Mol models, we recommend using a GPU for significantly faster training.

Notice: [Uni-Core](https://github.com/dptech-corp/Uni-Core) is needed, please install it first. Current Uni-Core requires torch>=2.0.0 by default, if you want to install other version, please check its [Installation Documentation](https://github.com/dptech-corp/Uni-Core#installation).


Uni-Mol is uses Python 3.8+ and all models are built with [PyTorch](https://pytorch.org/). See [Installation](#installation) for details on how to install Uni-Mol and its dependencies.