Blame for README.md

---
language:
  - tr
license: cc-by-4.0
pretty_name: Antalia evaluation suites and training-data manifests
task_categories:
  - text-to-speech
tags:
  - turkish
  - tts
  - evaluation
  - manifests
size_categories:
  - 100K<n<1M
configs:
  - config_name: evaluation
    data_files:
      - split: turkish_v2
        path: evaluation/turkish-v2.jsonl
      - split: turkish_v1
        path: evaluation/turkish-v1.jsonl
      - split: pronunciation_v1
        path: evaluation/turkish-pronunciation-v1.jsonl
  - config_name: common_voice_26_tr_filtered
    data_files:
      - split: train
        path: manifests/common-voice-26.0-tr-filtered.train.jsonl
      - split: validation
        path: manifests/common-voice-26.0-tr-filtered.validation.jsonl
      - split: test
        path: manifests/common-voice-26.0-tr-filtered.test.jsonl
      - split: rejected
        path: manifests/common-voice-26.0-tr-filtered.rejected.jsonl
  - config_name: fleurs_tr
    data_files:
      - split: all
        path: manifests/fleurs-tr.jsonl
---

# Antalia evaluation suites and training-data manifests

Companion data for [Antalia 1](https://huggingface.co/cloud0day3/antalia-1) and
[Antalia 1 Foundation](https://huggingface.co/cloud0day3/antalia-1-foundation), an open Turkish
text-to-speech release whose development is discontinued. **No audio is included.** The repository
contains the prompt suites we report numbers on, the manifests that reproduce our public-corpus
filtering, the native-listening (CMOS) protocol with its raw single-listener results, and the
aggregate statistics used by the best-of-N selector.

The speaker's own recordings are published separately as
[antalia-voice-corpus](https://huggingface.co/datasets/cloud0day3/antalia-voice-corpus) (5.008 h, CC-BY-4.0).

Code: https://github.com/0daycloud/antalia · Paper: https://github.com/0daycloud/antalia/blob/main/paper/main.pdf

## Contents

### `evaluation/` — prompt suites (CC-BY-4.0, written by the authors)

| File | Prompts | Notes |
|---|---:|---|
| `turkish-v2.jsonl` | 120 | 10 categories × 12: general, voice_agent, names_places, numeric, foreign_abbreviations, adversarial_normalization, questions_confirmations, acknowledgement, emotional_style, long_form. All reported CER/WER/similarity numbers use this suite. |
| `turkish-v1.jsonl` | 39 | Earlier suite used for the FastPitch baseline failure analysis. `ack-002` collides with two Common Voice sentences; neither was in any training manifest. |
| `turkish-pronunciation-v1.jsonl` | 39 | Pronunciation stress set (foreign names, abbreviations, numerals). |

Fields: `id`, `category`, `text`. Prompts were deduplicated against every training manifest.

### `manifests/` — public-corpus filter manifests

Reproduce exactly which upstream clips passed our gates. Audio is **not** re-hosted; obtain it from
Mozilla Common Voice 26.0 (Turkish) and Google FLEURS (`tr_tr`) under their own terms.

| File | Records | Meaning |
|---|---:|---|
| `common-voice-26.0-tr-filtered.train.jsonl` | 101,622 | Accepted clips, speaker-disjoint train split (110.2 h) |
| `common-voice-26.0-tr-filtered.validation.jsonl` | 5,014 | Accepted, validation split (5.1 h) |
| `common-voice-26.0-tr-filtered.test.jsonl` | 4,824 | Accepted, test split (5.1 h) |
| `common-voice-26.0-tr-filtered.rejected.jsonl` | 8,947 | Rejected clips with `quality_filter_reasons` |
| `common-voice-26.0-tr-filtered.report.json` | — | Aggregate counts and rejection reasons |
| `fleurs-tr.jsonl` | 3,607 | FLEURS Turkish clips (dataset revision `70bb2e84…`) |

Per record: `clip_id`, `source_file` (upstream file name, e.g. `common_voice_tr_29908728.mp3`),
`transcript`, `normalized_transcript`, `duration_seconds`, `sample_rate_hz`, `sha256` of the
upstream audio file, acoustic QA (`common_voice_acoustics`: peak, RMS dBFS, clipping ratio, active
frame ratio, estimated SNR), split, license id, attribution.

**Removed on purpose:** speaker identifiers, age, gender, vote counts, and local paths. The
Mozilla Data Collective terms ask users not to attempt speaker identification; these manifests
add no capability beyond the upstream dataset. The foundation model's training subset
(59,593 Common Voice + 1,876 FLEURS clips, 67.55 h) was drawn from the accepted train split with
per-speaker caps applied on upstream `client_id`; the cap procedure is in the code repository
(`src/turkish_tts/common_voice_prepare.py`).

### `listening/` — native-listener CMOS protocol and results

`cmos-protocol.md` (≥3-listener protocol as designed), `cmos-v1-instructions.md` (Turkish listener
sheet), `cmos-v1-trials.csv` (30 trials: 15 A/B gated-vs-plain, 12 anchored synth-vs-real, 3 catch),
`cmos-v1-key.json` (which system was A/B), `cmos-v1-results-listener-1.csv` (the only completed
session: one native listener, the project owner), `cmos-v1-results.md` (analysis: pooled CMOS
−0.667 ± 0.653; anchored −1.833 ± 0.757; same-person 0/12). The trial audio is not included because
half of the anchored pairs are real recordings of the voice actor.

### `selection/` — best-of-N selector inputs

`inference-recipe.json` (champion sampling recipe), `prosody-presets.json` (six style targets),
`timbre-profile.json` (level-normalized 11-band spectrum mean/std of 13 real recordings),
`envelope-stats.json` (F0/energy/rate statistics). These are aggregate statistics; they do not
contain audio.

## Licenses

Prompt suites, protocol documents, presets, and statistics: CC-BY-4.0 (attribution: the Antalia
authors). Manifests describe third-party data: Common Voice clips are CC0-1.0 under the Mozilla
Data Collective terms; FLEURS is CC-BY-4.0 (Google). The manifests themselves are released
CC-BY-4.0.

## Citation

```bibtex
@misc{antalia1_2026,
  title  = {Antalia 1: An Open Turkish Text-to-Speech Model from a Rights-Clean Pipeline},
  author = {Saygili, Sezgin and Kaplaner, Emre and Ozgul, Oncel and Koktas, Fikri San},
  year   = {2026},
  note   = {Technical report},
  url    = {https://huggingface.co/datasets/cloud0day3/antalia-eval}
}
```

Contact: sezgin@patientdesk.ai