This release contains only selected synthetic training samples, the frozen L1/L2 full/hard evaluation sets, and the full merged V3 L1 calibrated250 checkpoint. The publisher has confirmed permission to redistribute the included data. Original source rights and attribution remain applicable; no blanket replacement license is asserted. - releases/20261004/train/: independently extractable shards, each with manifest.jsonl, selected audio, source hashes, and SHA256SUMS.json. Union membership is specified by selectedforl1 and selectedforl2; do not count a shared sample twice. No real training samples, rejected candidate regions, failed generations, or feature caches are included.…
Independent publisher
CSY20060325
CSY20060325
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.
Published tasks: Audio classification.
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.