lightning/examples/fabric/kfold_cv

K-Fold Cross Validation

This is an example of performing K-Fold cross validation supported with Lightning Fabric. To learn more about cross validation, check out this article.

We use the MNIST dataset to train a simple CNN model. We create the k-fold cross validation splits using the ModelSelection.KFold class in the scikit-learn library. Ensure that you have the scikit-learn library installed;

pip install scikit-learn

Run K-Fold Image Classification with Lightning Fabric

This script shows you how to scale the pure PyTorch code to enable GPU and multi-GPU training using Lightning Fabric.

# CPU
lightning run model train_fabric.py

# GPU (CUDA or M1 Mac)
lightning run model train_fabric.py --accelerator=gpu

# Multiple GPUs
lightning run model train_fabric.py --accelerator=gpu --devices=4

References

• KFold Model Selection
• K-Fold Cross Validation by Sebastian Rashcka
• Cross Validation Wiki