SilencioNetwork/cebuano-speech

Cebuano (Bisaya) Spontaneous Speech — Silencio Philippines Pack

Spontaneous long-form Cebuano with human transcription and word-level forced alignment. Fifteen speakers, mean clip length over two minutes, 27,000+ timestamped tokens. Part of the Silencio Philippines Pack.

All 90 clips carry a human transcription.

Recordings are unscripted responses to open prompts, captured on contributors' own devices in their own environments. Mean clip length is 139 seconds — long-form spontaneous speech, not short read utterances.

This is a sample, not the catalogue. It is drawn to show format, audio quality and metadata, and its speaker mix does not represent what Silencio holds. The off-the-shelf catalogue behind it is far more diverse across regions, age groups, accents and recording conditions. Men are 60% of the speakers in this sample. Across the catalogue the gender split is roughly 40% female to 60% male, with men in the majority in most languages. Licensed subsets can be drawn to a specified gender, age or regional balance.

Load it

from datasets import load_dataset

ds = load_dataset("SilencioNetwork/cebuano-speech", split="train")
print(ds[0]["transcript"], ds[0]["dialect"], ds[0]["country"])

# datasets v4 returns a torchcodec AudioDecoder:
s = ds[0]["audio"].get_all_samples()
audio, sr = s.data, s.sample_rate

Requires pip install "datasets>=4.0" and FFmpeg ≥ 4.

Speaker and recording metadata

By country

Speaker origin / self-reported variety — this is the speaker's own background, not a dialect classification of the recorded language

Demographics

Recording conditions

Splits

Single split, test, 90 rows. No train/dev/test partition is provided: at this scale a partition would leave each part too small to be meaningful. Speaker identifiers are stable, so a speaker-disjoint split can be constructed at load time.

Fields

Related Cebuano and Philippine speech resources

Cebuano (Bisaya, Binisaya) has roughly 20 million speakers across the Central Visayas, Negros Oriental and much of Mindanao — the second most widely spoken language in the Philippines. Existing Hub coverage:

There is no single-language Cebuano audio dataset on the Hub. FLEURS is the closest usable resource and is read speech in short utterances; this release is unscripted long-form speech with per-word timings. The two are complementary rather than competing — FLEURS for read-speech benchmarking, this for spontaneous-speech behaviour.

Also from Silencio. Tagalog / Filipino is published under this same protocol — spontaneous speech, human transcription, word-level alignment. Hiligaynon and expanded Cebuano follow; see SilencioNetwork.

Transcription and alignment

Two distinct provenances, kept separate because they carry different confidence.

Text — human. Every transcript was produced by a human annotator listening to the recording. The transcript_model column records this per clip.

Timings — machine. Word-level start and end times come from forced alignment with wav2vec2_mms, recorded per clip in the aligner column. On every clip in this release the aligner's token count matches the human reference token count exactly, and no word timing runs past the end of its audio file.

The words column holds one entry per token with text, normalized_text, start and end in seconds. Expand it for segment-level work:

ds = load_dataset("SilencioNetwork/cebuano-speech", split="test")
row = ds[0]
for w in row["words"][:5]:
    print(f"{w['start']:6.2f}-{w['end']:6.2f}  {w['text']}")

Speaker proficiency

Cebuano proficiency is taken from each contributor's own declared language profile, not inferred from a single primary-language field. Most contributors here are natively bilingual: their primary declared language is Tagalog, and they also declare Cebuano at native level.

Filter on native_speaker, or on proficiency for finer control.

What this is useful for

• —Spontaneous-speech ASR evaluation. Long-form unscripted Cebuano with human reference text. Models tuned on read speech typically degrade sharply here; that gap is the point.
• —Forced-alignment and VAD work. 27,000+ word-level timings over 3.5 hours.
• —Long-form segmentation. Mean clip length 139 s, with several clips over four minutes. Most Philippine-language audio on the Hub is short read utterances.
• —Code-switching and borrowing. Spontaneous Cebuano from bilingual speakers contains substantial Tagalog, Spanish and English material. Not annotated as such in this release.

Limitations

• —See the demographic tables above for balance across gender, age and region.

Provenance and consent

Every recording is contributed by an opted-in participant through the Silencio platform, under a consent record covering AI/ML training use. Contributors can request deletion, and deletion propagates to downstream releases. Full provenance documentation is available to licensees.

License

cc-by-nc-4.0 — free for research and non-commercial use with attribution.

Attribution string: Silencio Network, Cebuano (Bisaya) Spontaneous Speech, 2026. CC BY-NC 4.0.

Non-commercial covers research, evaluation and publication. Benchmarking a commercial product model against this data is a commercial use and needs a licence — ask, it is usually granted for evaluation. Model weights trained on this sample inherit the non-commercial restriction. Contributors may withdraw consent; withdrawal propagates to subsequent releases but places no retroactive obligation on an existing licensee.

Commercial licensing, including terms for models trained on this data: [silencio.network/contact](https://www.silencio.network/contact)

Citation

@misc{silencio_cebuano_2026,
  title  = {Cebuano (Bisaya) Spontaneous Speech — Silencio Philippines Pack},
  author = {Silencio Network},
  year   = {2026},
  url    = {https://huggingface.co/datasets/SilencioNetwork/cebuano-speech}
}

The Silencio Philippines Pack

Philippine catalogue depth, September 2026, shown as bands:

In active collection: 7,500 hours. A collection and human-transcription programme covering 2,500 hours each of Cebuano, Tagalog and Hiligaynon, split per language into 1,000 hours single-speaker and 1,500 hours multi-speaker.

Human-validated transcription is available for every language, scoped to the client's needs, and every clip in delivery is validated by a native-speaker human reviewer.

Ilocano, Waray, Bikol, Kapampangan and Pangasinan are available through commissioned collection.

Silencio: the catalogue behind this sample

This listing is a small fraction of what Silencio provides. Silencio specialises in underrepresented languages, accents and niche domains, delivered at scale. It draws on the largest global community of speech contributors and holds the largest off-the-shelf catalogue of niche-language speech available.

Recorded and available off the shelf. Audio already collected, with metadata, licensable today. Catalogue position as of September 2026:

Contributor community available on demand. Registered, consented contributors who can be activated for a specific brief:

Anything not off the shelf can be sourced through the community. A language, an accent or dialect region, a demographic, a recording condition, a speech style or a specialist domain can be collected to a client's brief. Deep coverage across Africa, South-East Asia, South Asia and the Middle East.

Human-validated transcription is available for every language, scoped to each client's needs: script and orthography conventions, normalisation rules, word- or segment-level alignment, speaker labelling and turnaround. Every delivered clip is checked by a native-speaker reviewer. Published samples with human-validated transcripts: Kenyan Swahili, Cebuano, Tagalog / Filipino, Yoruba, Hausa and Amharic.

Proprietary and first-party. Every recording is collected directly by Silencio from consenting contributors. Nothing is scraped, and these recordings are not available in any other dataset on the internet.

Ethical sourcing, with provenance records. Every recording carries a consent record covering AI/ML training use, and contributor-level provenance documentation is available to licensees, including for EU AI Act training-data summaries. Contributors can withdraw consent, and withdrawal propagates to subsequent releases.

More at silencio.network.

For volume licensing, bespoke transcription or commissioned collection: [silencio.network/contact](https://www.silencio.network/contact)