The dataset viewer is not available for this subset.
Exception: SplitsNotFoundError
Message: The split names could not be parsed from the dataset config.
Traceback: Traceback (most recent call last):
File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 286, in get_dataset_config_info
for split_generator in builder._split_generators(
~~~~~~~~~~~~~~~~~~~~~~~~~^
StreamingDownloadManager(base_path=builder.base_path, download_config=download_config)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
)
^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/folder_based_builder/folder_based_builder.py", line 246, in _split_generators
raise ValueError(
"`file_name`, `*_file_name`, `file_names` or `*_file_names` must be present as dictionary key in metadata files"
)
ValueError: `file_name`, `*_file_name`, `file_names` or `*_file_names` must be present as dictionary key in metadata files
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "/src/services/worker/src/worker/job_runners/config/split_names.py", line 71, in compute_split_names_from_streaming_response
for split in get_dataset_split_names(
~~~~~~~~~~~~~~~~~~~~~~~^
path=dataset,
^^^^^^^^^^^^^
config_name=config,
^^^^^^^^^^^^^^^^^^^
token=hf_token,
^^^^^^^^^^^^^^^
)
^
File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 340, in get_dataset_split_names
info = get_dataset_config_info(
path,
...<6 lines>...
**config_kwargs,
)
File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 291, in get_dataset_config_info
raise SplitsNotFoundError("The split names could not be parsed from the dataset config.") from err
datasets.inspect.SplitsNotFoundError: The split names could not be parsed from the dataset config.Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
Sonora Expressive Registers
Labeled expressive speech clips for training and evaluating directable TTS — the companion dataset of the Sonora model line (Project Prosodia, Artificial Humanity).
Openly-licensed expressive speech is scarce: most emotion-labeled corpora are non-commercial-encumbered. This dataset is CC-BY-4.0 by construction: every clip is synthesized by Apache/MIT-licensed teacher engines from in-repo authored or public-domain text, with exact intended emotion labels (the render direction is the label) — no consent surface, no license taint.
What a clip carries (metadata.jsonl)
| Field | Meaning |
|---|---|
register |
expressive register label (controlled, owner-governed lexicon) |
intended_vat |
the directed Valence / Arousal / Tension in [-1, 1] — exact by construction |
measured_z |
instrument measurements on the rendered audio (loudness, phonation-composite tension, arousal) — the cross-check |
direction |
the full render direction (voice design + delivery instruction) |
engine, seed |
teacher engine + seed — every clip is reproducible |
qc |
automated gate results (ASR/WER, DNSMOS, duration sanity) |
owner_audit / owner_relabel / gender |
human audition verdict, relabels, audited speaker gender |
text, text_source, license |
the spoken text and its provenance |
Certification bar: a clip enters a release only after passing the automated QC gate and a human blind audition ("affect obvious without the keyword"). Labels the audit contradicted are relabeled or dropped — audited values are the source of truth.
Composition (v1)
193 clips · ~66 MB · 24 kHz mono WAV · single-utterance (1–16 s) · English. Registers span the expressive space (arrogance, grief, fear-of-loss, tenderness, menace, triumph, …) rendered by a multi-engine teacher portfolio (Dia / Qwen3-TTS VoiceDesign / MOSS-TTSD) under an LLM emotional-director pass.
Own-synthesis only, by policy: this dataset never includes clips derived from external audio sources (LibriVox, LibriTTS-R, etc.) — only our own directed synthesis. Versioned per campaign; grows continuously as new campaigns certify.
Intended uses
Training/fine-tuning expressive or emotion-conditioned TTS; evaluating emotion conveyance;
prosody research. The intended_vat ↔ measured_z pairing also supports studying how
directed emotion maps to acoustics.
Provenance
Full pipeline documentation ships in PROVENANCE.md. Synthesis engines: Apache-2.0/MIT.
Director: Gemma 4 (Apache-2.0). Texts: authored in-project or public domain.
- Downloads last month
- 251