Datasetpublic

zuhri025/Urdu-Munch-Curriculum-easy

Urdu Munch Curriculum Easy Filtered Urdu TTS training subset for easier, shorter audio sequences. Split train: 242549 rows Curriculum Rule audio_token_len < 120 Fields id transcript voice text timestamp duration audio_content_token_indices audio_global_embedding audio_token_len Notes This dataset was created by filtering zuhri025/Urdu-Munch-Processed-s-merged for curriculum learning.

sourceHugging Faceupdated 5mo agoView on Hugging Face

0likes12downloads

Dataset Card

Urdu Munch Curriculum Easy

Filtered Urdu TTS training subset for easier, shorter audio sequences.

Split

—train: 242549 rows

Curriculum Rule

—audiotokenlen < 120

Fields

—id
—transcript
—voice
—text
—timestamp
—duration
—audio_content_token_indices
—audio_global_embedding
—audio_token_len

Notes

This dataset was created by filtering zuhri025/Urdu-Munch-Processed-s-merged for curriculum learning.