Implementation:Huggingface Datasets Dataset Num Rows

Knowledge Sources	Huggingface Datasets HF Datasets Docs
Domains	Data_Engineering, ML_Preprocessing
Last Updated	2026-02-14 18:00 GMT

Overview

Concrete tool for checking the number of rows in a dataset provided by the HuggingFace Datasets library.

Description

The num_rows property returns the total number of rows (examples) in the dataset as an integer. If the dataset has an indices mapping (e.g., after shuffling or selecting), it returns the number of mapped indices. Otherwise, it returns the number of rows in the underlying Arrow table. This property is equivalent to calling len(dataset).

Usage

Use Dataset.num_rows when you need to know the size of a dataset for computing split proportions, setting batch sizes, validating data loading, or reporting dataset statistics.

Code Reference

Source Location

• Repository: datasets
• File: src/datasets/arrow_dataset.py
• Lines: L1879-L1894

Signature

@property
def num_rows(self) -> int:

Import

from datasets import load_dataset

ds = load_dataset("cornell-movie-review-data/rotten_tomatoes", split="validation")
ds.num_rows

I/O Contract

Inputs

Name	Type	Required	Description
(none)	N/A	N/A	This is a property with no parameters.

Outputs

Name	Type	Description
return	int	The number of rows in the dataset. Returns the indices count if an indices mapping exists, otherwise the Arrow table row count.

Usage Examples

Basic Usage

from datasets import load_dataset

ds = load_dataset("cornell-movie-review-data/rotten_tomatoes", split="validation")
print(ds.num_rows)
# 1066

Related Pages

Implements Principle

• Principle:Huggingface_Datasets_Row_Count_Inspection