GLM-5.2 is live. #1 throughput on OpenRouter, pay-per-token on FriendliAI. Try it today ➜
Slavik2009
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Run this model inference with full control and performance in your environment.
Model Details
Model Provider
Slavik2009
Model Tree
openai/whisper-large-v3-turbo
Input Modalities
Output Modalities
Supported Functionality
| Model | Task |
|---|---|
| TurkmenChat-3B | 💬 Conversational LLM |
| whisper-large-v3-turbo-tk | 🎙️ Speech-to-text |
every session — all numbers below are measured on the SAME data)
| Model | WER, % | CER, % |
|---|---|---|
| Whisper large-v3-turbo, zero-shot | 117.93 | 72.45 |
| This model (fine-tuned) | 34.74 | 7.30 |
| Improvement | -83.2 pts |
| Session | WER, % | CER, % |
|---|---|---|
| 2 | 39.75 | 8.60 |
| 3 | 34.74 | 7.30 |
Best WER so far: 34.74% | Sessions: 3 | Corpus position: 29250 files | This session: 8000 train clips
| Reference (human) | Base Whisper (zero-shot) | This model |
|---|---|---|
| esalawmalikim hürmeti adamlar garaşsyz baky bitarap ýurdumyzdaky | Assalamualaikum warimata adammar garas shada bachibi tarabi yorna madhaqe | sulumalikim hurmatly adamlar garaşşada baky bitarap ýurdumyzdaky |
| dürdäne sygyrlaryny içgin öwrenýärler | Durdane-sağırlar, nechkine oran yarlarlar, | dürdäne sygyrlaryny içgine öwrenýärler |
| öz dillerine terjime edýärler | ... | öz dillerine terjime etýärler |
| çünki arkadagly gahrymançy ýerdarymyzyň | Tшоң кей, арқыда алға қалманшердармыдың | çünki arkadagly gahlyman serdarymyzyň |
| bu ajaýyp kitabyndade bellenişi ýaly | Bu acayip kıtahmla de bellensi ali, |
Full comparison data: benchmark_samples.json in this repo.
python
from transformers import pipelineasr = pipeline("automatic-speech-recognition", model="Slavik2009/whisper-large-v3-turbo-tk")print(asr("audio.mp3"))
Transcribes Latin-script spoken Turkmen (broadcast/podcast-style audio, 1-30s clips at 16 kHz). Quality degrades on heavy dialects, far-field or noisy audio, and code-switched speech. WER/CER above are measured on a held-out slice of the training corpus, not an independent test set — treat them as an upper bound of in-domain quality.
bibtex
@misc{whisperturbotk,title = {whisper-large-v3-turbo-tk: Whisper large-v3-turbo fine-tuned for Turkmen ASR},author = {Slavik2009},year = {2026},url = {https://huggingface.co/Slavik2009/whisper-large-v3-turbo-tk}}
| bu ajaýyp kitabynda-da bellenişiýaly |