Vekol STT (ckb) Edge
Copyright (c) 2026 Darvan Shvan

This repository contains code and fine-tuned model weights for Central Kurdish
(Sorani) speech-to-text, released under CC-BY-NC 4.0 (see LICENSE).

Attribution
-----------
- Base models: OpenAI Whisper (tiny/base/small), MIT License.
  https://github.com/openai/whisper
- Training data:
  - Mozilla Common Voice 25.0 (ckb), CC0.
  - Google FLEURS (ckb_iq), CC-BY 4.0.

The fine-tuned weights in this repository are derivative works of the MIT-licensed
Whisper models. The weights and the code here are distributed under CC-BY-NC 4.0.

For commercial use, see the hosted service at https://vekol.krd or contact the authors.