Skip to content

Dataset · Hebrish

Single-speaker code-switching English and Hebrew voice recordings

Hebrish/Hebrish-TTS-Voice-Recordings

Speech recordings of a single speaker with an Irish accent mixing English and Hebrew for speech synthesis training.

What's inside

A standalone audio dataset of 10k to 100k records containing audio and text in English and Hebrew. It features both read and spontaneous speech from a single speaker, Daniel Rosehill, who lives in Israel and speaks with an Irish accent. The data reflects natural immigrant code-switching ("Hebrish") and is designed for text-to-speech tasks.

What you can do with it

  • Training a TTS model capable of code-switching between English and Hebrew in a single phrase
  • Cloning a speaker's voice with a specific accent
  • Testing speech synthesis on mixed vocabulary
Tasks
Text → speech
Languages
English, Hebrew
Size
10K–100K rows
Downloads this month
70.5K
Likes
0
Access
on request on Hugging Face

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("Hebrish/Hebrish-TTS-Voice-Recordings")