X Voice Dataset Train: A Multilingual Speech Corpus for Model Training

XRXRX aggregated this multilingual speech dataset from seven distinct sources, including Multilingual LibriSpeech, VoxPopuli, and GigaSpeech 2. The collection was last updated on April 12, 2026. Each constituent dataset retains its own license, with most permitting commercial use.

Use Cases

• Train multilingual automatic speech recognition (ASR) models based on the described multilingual sources.
• Fine‑tune text‑to‑speech (TTS) systems using the aggregated voice data.
• Develop speech synthesis models leveraging the combined audio corpus.
• Benchmark speech model performance across different languages and accents represented in the source datasets.

Strengths

• Aggregates data from seven established speech datasets, providing a broad base.
• Most listed source datasets explicitly permit commercial use under licenses like CC BY 4.0.
• The dataset page was updated on 2026‑04‑12, suggesting recent maintenance.

Limitations

• Column‑level documentation is absent; field semantics must be inferred after download.
• Row count and total size are unknown, which may limit suitability assessment.
• License terms are not uniform; users must verify compliance for each sub‑dataset used.

Provenance

Source

Aggregated from seven sources: Multilingual LibriSpeech, Emilia, LEMAS, VoxPopuli, Granary (MOSEL Part), GigaSpeech 2, and Reazon Speech.

Collection Method

Likely a curated compilation of existing public speech datasets.

Freshness

Last updated 2026‑04‑12 16:31:54; freshness should be verified.