| --- |
| license: apache-2.0 |
| language: |
| - kk |
| base_model: |
| - microsoft/VibeVoice-ASR |
| tags: |
| - automatic-speech-recognition |
| - kazakh |
| - vibevoice |
| - lora |
| - ksc2 |
| datasets: |
| - InflexionLab/ISSAI-KSC2-Structured |
| metrics: |
| model-index: |
| - name: VibeVoice-ASR-Kazakh |
| results: |
| - task: |
| type: ASR |
| dataset: |
| type: AudioDataset |
| name: ISSAI KSC2 |
| metrics: |
| - type: WER |
| value: 22% |
| --- |
| |
|
|
| # VibeVoice ASR — Kazakh |
|
|
| ## Model Description |
|
|
| This is **VibeVoice ASR** fine-tuned on the Kazakh language using the [ISSAI KSC2 |
| Structured](https://huggingface.co/datasets/InflexionLab/ISSAI-KSC2-Structured) dataset (~1,200 hours of diverse Kazakh speech). Fine-tuning was performed |
| using LoRA (Low-Rank Adaptation) and the weights were merged into the base model for efficient inference. Model demonstrated **22%** WER on test set of ISSAI KSC2. |
|
|
| The base VibeVoice ASR model had no prior Kazakh knowledge. This fine-tuned version produces punctuated and capitalized Kazakh transcriptions. |
|
|
| ## Training Dataset |
|
|
| [InflexionLab/ISSAI-KSC2-Structured](https://huggingface.co/datasets/InflexionLab/ISSAI-KSC2-Structured) — an enhanced version of the ISSAI KSC2 corpus |
| with punctuation and capitalization restored using Gemma 27B. Covers 6 domains: TV News, Crowdsourced, Parliament, Talkshow, Podcasts, and Radio. |
|
|
| ## Evaluation Results |
|
|
| Evaluated on the KSC2 Test split (9,351 samples) and farabi-lab/kazakh-stt 30K samples. Farabi-Lab dataset was not included in training. |
|
|
| | Dataset | WER | CER | |
| |----------------------------|-----------------|-----------------| |
| | ISSAI_KSC2 | ~22% | ~9.6% | |
| | farabi-lab/kazakh-stt | 17.6% | 4.25% | |
