rokeya71's picture
Update README.md
70e07fd verified
metadata
license: mit
language:
  - en

Everyday Conversations Fine-Tuning Dataset (LLaMA 3.1 - 2K)

Overview

This repository hosts the Everyday Conversations - LLaMA 3.1 - 2K dataset, a carefully curated fine-tuning dataset designed for conversational AI models. The dataset was created using the Kokoro-82M model, featuring voice samples from the af Voicepack.

Dataset Link

Features

  • Voice Model: Kokoro-82M
  • Voicepack: af (American, Female)
  • Samples: 2,000 24kHz conversational text and audio pairs
  • Language: English (American accent)