One config per corpus, each a set of parquet shards whose rows carry the audio of every segment of one pretraining record, so a corpus can be streamed front to back: Rows: id, source, seconds, numsegments, meta (JSON), and segments, a (16-bit FLAC unless the corpus's packinfo.json says otherwise), mono 16 kHz. Rows are pre-shuffled; each corpus has a multiple of 8 shards. Each corpus keeps its own licence -- see /packinfo.json. Built by scripts/packstreaming.py.
Publisher
Giahy2507
giahy2507
Models in Library0
Datasets in Library1
Models on Hugging Face—
Followers—
Compare the published evidence
This profile connects the 1 model and dataset record shown below to their upstream cards and sources. The library counts describe SAVRN's coverage, rather than this publisher's entire output.
Compare model specifications, source dates, and hardware estimates on the individual records. Publisher identity verification is separate from a model's quality, deployment readiness, or suitability for your work.