You need to agree to share your Hugging Face information to access this dataset

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Log in or Sign Up to review the conditions and access this dataset content.

multi_ds_qezel_arrow

Japanese podcast speech as Fish codec features (latents float16 (T, 80) at ~21.6 frames/s, codes uint16 (10, T)) with clip metadata; no audio and no transcripts. Each config folder is a datasets.save_to_disk layout:

snapshot_download(repo, repo_type='dataset', allow_patterns='<config>/*', local_dir='x')
ds = datasets.load_from_disk('x/<config>')   # memory-mapped, no conversion
Downloads last month
253