Antalia 1 Foundation — speaker-agnostic Turkish acoustic model

The base model behind Antalia 1: a ~300M-parameter character-conditioned rectified-flow model over 24 kHz, 100-band log-mel spectrograms, trained from scratch on filtered public Turkish corpora only. It has no speaker conditioning; the voice of any output is an unspecified average of Common Voice contributors and varies with the sampling seed. It is released so that others can fine-tune their own consented Turkish voice with the code in https://github.com/0daycloud/antalia.

Development is discontinued. Released as-is.

Code: https://github.com/0daycloud/antalia (Apache-2.0)
Fine-tuned voice: https://huggingface.co/cloud0day3/antalia-1
Voice corpus used for that fine-tune (CC-BY-4.0): https://huggingface.co/datasets/cloud0day3/antalia-voice-corpus
Paper: https://github.com/0daycloud/antalia/blob/main/paper/main.pdf
Contact: sezgin@patientdesk.ai or GitHub issues
Türkçe: README.tr.md

Training data

Source	Clips used	License
Common Voice 26.0 Turkish (validated, filtered)	59,593	CC0-1.0
FLEURS Turkish	1,876	CC-BY-4.0

Total 61,469 clips / 67.55 h. Filtering: forced-Turkish Whisper-large-v3 transcript check (accepted median CER 0, p90 0.111), active-audio, level, clipping, and SNR gates, per-speaker caps, speaker-disjoint validation split. Rejections from 120,407 validated CV clips: 7,098 insufficient active audio, 1,449 transcript mismatch, 356 too short, 351 low level, 5 low SNR, 4 clipping. Filter manifests with clip ids and checksums ship in the code repository; the audio itself is not re-hosted.

Training

Clean initialization; 100,000 updates on one A100-80GB; AdamW, LR 1e-4 with 2,000 warmup updates decaying to 1e-5, weight decay 0.01, gradient clip 1.0, EMA 0.9999; 6,000 mel frames per batch (≤32 samples, ≤24 s per clip); duration-loss weight 0.1; per-band mel mean/std normalization measured on 256 training records (89,966 frames), clipped at ±5. Two earlier foundation runs without per-band normalization diverged and were abandoned.

Quality

On the 120-prompt turkish-v2 suite, single seed, 32 Euler steps, no guidance: CER 0.1835 (p90 0.4049), WER 0.3952. This is an intelligibility baseline for a speaker-agnostic model with a sampled voice; it is not usable as a product voice on its own. The fine-tuned Antalia 1 reaches CER 0.0528 single-seed on the same suite.

Files

File	Purpose
`model.safetensors`	EMA weights, fp32, 299,623,013 parameters. SHA-256 `89ecf310a14333c6bd5a754360cee2207027dd23b2a0171984a32c423afcfcd1`
`config.json`	Release format `antalia-crossflow-release-v1`: architecture, mel statistics, character vocabulary, vocoder pointer, provenance

Vocoder: nvidia/bigvgan_v2_24khz_100band_256x (MIT), downloaded by the loader.

How to run

python scripts/synthesize-crossflow.py \
  --checkpoint cloud0day3/antalia-1-foundation \
  --vocoder nvidia/bigvgan_v2_24khz_100band_256x \
  --sway -0.8 --steps 32 --mel-clamp 5.0 \
  --text "Bugün hava çok güzel, dışarıda yürüyüş yapmak istiyorum." \
  --output foundation.wav

To fine-tune a new voice, see the training configs configs/crossflow/foundation-speaker-v2.json → candidate-b-speaker-v1.json → cfg-foundation-v1.json → candidate-consistency-v1.json → candidate-b-timbre-adapter-v2.json in the repository as the reference lineage.

Architecture

16 Transformer blocks (dim 768, 12 heads, SwiGLU FFN 3072) with self-attention over mel frames, cross-attention over characters, and zero-initialized adaLN modulation from the flow timestep; 4-block ConvNeXt-style character encoder; a scalar log-total-duration head. Deterministic Turkish text normalizer; grapheme input.

Limitations

Unconditioned voice identity is arbitrary and unstable across seeds. Long inputs are under-budgeted by the duration head (use clause chunking). Numbers, foreign names, and abbreviations have the highest error rates. Turkish only. No memorization audit was run.

License and use

Weights: Antalia Open RAIL-M (see LICENSE.md): no impersonation, no undisclosed synthetic speech, no fraud or robocalls. Code: Apache-2.0.

Citation

@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/cloud0day3/antalia-1-foundation}
}

Downloads last month: -

Safetensors

Model size

0.3B params

Tensor type

F32

Model tree for cloud0day3/antalia-1-foundation

Finetunes

1 model

cloud0day3
/

antalia-1-foundation