What is k-fold cross-validation?

K-fold cross-validation splits a dataset into k equally-sized folds. The model is trained k times, each time using k-1 folds for training and 1 fold for validation. This gives a more reliable performance estimate than a single train/test split.

How are remainder samples handled?

When the dataset size is not evenly divisible by k, the remainder r = dataset_size mod k is distributed by making the first r folds one sample larger than the rest. This is the standard behavior in scikit-learn's KFold.

What is a good value for k?

k=5 and k=10 are the most common choices. Larger k gives a larger training set per fold but increases compute time. k equal to the dataset size is called leave-one-out cross-validation (LOOCV).

Does fold order matter?

The fold breakdown shown here assumes data is already shuffled. If your data has temporal ordering or groups, use stratified or grouped k-fold instead, but the size calculations remain the same.

Why does each fold have a different validation size?

Only the first (dataset_size mod k) folds get one extra sample to absorb the remainder. All other folds have the base size (floor(dataset_size / k)).

K-Fold Cross-Validation Size Calculator

Enter your dataset size and number of folds to see exact training and validation sample counts for each fold.

Dataset Parameters

Dataset Size (N)

Total number of samples in your dataset.

Number of Folds (k)

Typical values: 5 or 10.

Fold Breakdown

Enter parameters and click Calculate Splits to see fold breakdown.

Fold	Validation samples	Training samples	Val %	Train %

Copied to clipboard

Summary

Enter your dataset size and number of folds to see exact training and validation sample counts for each fold.

How it works

Enter the total number of samples in your dataset.
Set k — the number of folds you want to split the data into.
The calculator divides the dataset into k roughly equal folds.
Remainder samples (when dataset size is not divisible by k) are distributed one per fold across the first folds.
Each row shows which fold acts as the validation set while the rest form the training set.
Use the results to allocate data in your cross-validation loop.

Use cases

Verify fold sizes before writing a cross-validation training loop.
Plan compute time by knowing exact training set sizes per fold.
Check how remainder samples are distributed when dataset size is not divisible by k.
Compare the effect of different k values on validation set size.
Communicate dataset splits clearly in research papers or team discussions.
Quickly estimate validation set size for a given dataset and fold count.

Frequently Asked Questions

Last updated: 2026-09-19 · Reviewed by Nham Vu

K-Fold Cross-Validation Size Calculator

Use the K-Fold Cross-Validation Size Calculator

Dataset Parameters

Summary

Fold Breakdown

Fold size visualization

Summary

How it works

Use cases

Frequently Asked Questions