---
language:
- en
task_categories:
- text-generation
tags:
- names
- pronunciation
- variants
pretty_name: Sheet 2 - grouped name variants
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-00000-of-00001.parquet
---

# Sheet 2: grouped name variants

Audio-free grouped pronunciation data derived only from approved Sheet 1
names. Each row is one approved name with multiple observed pronunciations;
`pronunciations` preserves every phonetic spelling, mapped IPA, and probability.

This snapshot contains 300 approved names and 895 grouped pronunciation
variants.