You need to agree to share your contact information to access this dataset

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Log in or Sign Up to review the conditions and access this dataset content.

Vietnamese F5-TTS Dataset

Mô tả Dataset

Dataset tiếng Việt được chuẩn bị cho việc fine-tuning mô hình F5-TTS. Dataset bao gồm 850909 mẫu audio-text pairs với chất lượng cao.

Thông tin Dataset

  • Ngôn ngữ: Tiếng Việt
  • Số lượng mẫu: 850909
  • Sample rate: 24kHz
  • Định dạng audio: WAV
  • Mục đích: Fine-tuning F5-TTS cho tiếng Việt

Cấu trúc Dataset

- audio: file âm thanh (WAV, 24kHz)
- text: nội dung văn bản tương ứng
- audio_path: đường dẫn file audio gốc

Cách sử dụng

from datasets import load_dataset

# Tải dataset
dataset = load_dataset("tuankg1028/vietvoices-tts-dataset")

# Truy cập một mẫu
sample = dataset["train"][0]
print(f"Text: {sample['text']}")
print(f"Audio: {sample['audio']}")

Fine-tuning với F5-TTS

Dataset này được thiết kế để sử dụng với F5-TTS cho tiếng Việt.

Giấy phép

MIT License - Xem file LICENSE để biết thêm chi tiết.

Đóng góp

Nếu bạn muốn đóng góp hoặc báo cáo lỗi, vui lòng tạo issue hoặc pull request.

Downloads last month
17