Access is gated because this repository contains speech-derived training artifacts. Request access only for approved training, evaluation, or research use.

By requesting access, you agree not to redistribute files, attempt speaker re-identification, or use the data for unauthorized voice impersonation.

tbp_chunked_speech_restorised_tts_train_clone_pairs

This is a gated Russian TTS training clone-pair dataset. It contains tokenized speaker-reference and target pairs for text-to-speech voice adaptation workflows.

Language

Primary language: ru (Russian)

audios/shard-*.tar: tokenized audio shards
txts/shard-*.jsonl: per-example metadata and text fields
data.lst: repository-relative shard manifest
tokenized_dataset_summary.json: upload-time dataset summary

Dataset Summary

Field	Value
Source dataset key	`tbp_chunked_speech_restorised`
Examples	42778
Duration	208.42 hours
Shards	43

Each data.lst row has this format:

<audio_tar> <metadata_jsonl> <examples> <seconds>

Intended Use

text-to-speech training
speaker-reference conditioning experiments
Internal dataset curation, quality checks, and evaluation
Research or commercial use only after access approval and license review

Data Notes

Prepared for TTS training workflows.
The examples are paired as reference speech plus target speech from the same source speaker grouping.
These are tokenized training artifacts, not a general-purpose audio playback dataset.

Access

This repository is gated with manual approval. Access requests are reviewed by the dataset maintainers. Users requesting access must provide their affiliation, intended use, and agreement to the dataset terms.