RidheshBhati/Complete_Data_Source_100K_HOURS

Multi-Language Audio Collection (100K Hours)

This repository hosts a massive 100,000+ hour audio dataset across 22 languages, physically reorganized for Absolute 100% Data Visibility.

🏗️ Partitioned Root Structure (V26)

The repository has been flattened into 22 professional language folders. To bypass the "10,000 files per directory" limit on the Hugging Face Hub, the data is stored in partitioned sub-folders (e.g., english/part_1/).

• Zero Clutter: Root folders data/ and shards/ have been eliminated.
• Total Searchability: All shards are now directly accessible through their language-specific paths.
• Audio Ready: Triple-Shield metadata ensures that all columns (audio, bytes, audio_bytes) render as playable audio in the Dataset Viewer.

🛡️ Omni-Router V5

This repository implements Omni-Router V5, which dynamically scouts all partitions and automatically sorts mixed shards into their correct linguistic configurations.

🌎 Global Language Support

All data is organized into 22 professional language configurations. Select your language from the subset dropdown to listen to audio and view transcripts.