CoolFace
Datasetpublic

PalakEngineerMaster/Processed_TTS_Multilingual_Data

Processed TTS Multilingual Data Validated and quality-checked multilingual speech datasets for TTS training, covering 12+ Indian languages. Datasets Included Subset Samples Hours Description indic_voices_r 239,684 548.8h Indic Voices_R — IVR recordings rasa 201,509 361.2h RASA — read speech (wiki, conv, book, news) indictts_iitm 155,236 253.6h Indic TTS (IIT Madras) — studio TTS recordings at 48kHz Total 596,429 1,163.6h… See the full description on the dataset page: https://huggingface.co/datasets/PalakEngineerMaster/Processed_TTS_Multilingual_Data.

sourceHugging Facecc-by-4.0updated 7mo agoView on Hugging Face
0likes6downloads
Dataset Card

Processed TTS Multilingual Data

Validated and quality-checked multilingual speech datasets for TTS training, covering 12+ Indian languages.

Datasets Included

SubsetSamplesHoursDescription
indic_voices_r239,684548.8hIndic Voices_R — IVR recordings
rasa201,509361.2hRASA — read speech (wiki, conv, book, news)
indictts_iitm155,236253.6hIndic TTS (IIT Madras) — studio TTS recordings at 48kHz
Total596,4291,163.6h

Languages

Assamese (as), Bengali (bn), English (en), Gujarati (gu), Hindi (hi), Kannada (kn), Malayalam (ml), Marathi (mr), Nepali (ne), Odia (or), Punjabi (pa), Tamil (ta), Telugu (te)

Structure

├── indic_voices_r/
│   ├── metadata.csv
│   └── audio/{lang}/*.wav
├── rasa/
│   ├── metadata.csv
│   └── audio/{lang}/*.wav
└── indictts_iitm/
    ├── metadata.csv
    └── audio/{lang}/*.wav

Schema (metadata.csv)

Each subset has a metadata.csv with these columns:

FieldDescription
file_nameRelative path to audio file (e.g., audio/bn/file.wav)
textTranscript text
langISO 639-1 language code
speaker_idSpeaker identifier
durationAudio duration in seconds
sourceOriginal data source
emotionEmotion label
domainText domain (wiki, conv, book, news, etc.)
snr_dbSignal-to-noise ratio in dB
silence_ratioFraction of silent frames
clipping_ratioFraction of clipped samples

Quality Checks Applied

All data has been validated through a 4-check pipeline:

  1. 1.SNR + Silence + Duration — reject low SNR (<10dB), excess silence (>35%), out-of-range duration (<1.5s or >30s), clipping (>1%)
  2. 2.Speaking Rate — reject abnormal speaking rates (<2 or >25 chars/sec)
  3. 3.Text Normalization — Unicode NFC normalization applied
  4. 4.Audio Corruption — reject empty, all-zeros, NaN/Inf, DC offset >0.1

Usage

python
from datasets import load_dataset

# Load a specific subset
ds = load_dataset(
    "PalakEngineerMaster/Processed_TTS_Multilingual_Data",
    data_dir="rasa",
    split="train",
)

# Access a sample
sample = ds[0]
print(sample["text"])
# audio is at sample["file_name"]

Audio Format

  • —Format: WAV
  • —Sample rate: 16kHz (Indic Voices_R, RASA) / 48kHz (Indic TTS IIT M)
  • —Channels: mono