You can deploy this model and start earning money today!

nicman23/
whisper-large-v3-turbo-greek-gguf

Name: whisper-large-v3-turbo-greek-gguf
Availability: InStock

Views

No views yet

Whisper Large V3 Turbo - Greek Finetuned (GGUF)

Greek-finetuned version of OpenAI's Whisper Large V3 Turbo model, converted to GGUF format for use with whisper.cpp and compatible inference engines.

Model Details

Base Model: openai/whisper-large-v3-turbo
Finetuned by: sam8000
Format: GGUF (whisper.cpp)
Quantization: Q5_K_M (5-bit K-quants)
File Size: 1.6 GB
Languages: Greek (el), English (en), +90+ other languages

Quantization

This model is quantized to Q5_K_M format:

Near-lossless accuracy compared to FP16 (~99%)
50% size reduction vs FP16 (3.2GB → 1.6GB)
Optimized for CPU/GPU inference
Recommended for production use

Usage

whisper.cpp

bash

# Clone whisper.cpp
git clone https://github.com/ggerganov/whisper.cpp
cd whisper.cpp
make

# Download model
./models/download-ggml-model.sh greek-large-v3-turbo

# Or use this file directly
./main -m /path/to/ggml-whisper-large-v3-turbo-greek-Q5_K_M.bin -f audio.wav -l el --no-timestamps

With Greek Language

./main -m ggml-whisper-large-v3-turbo-greek-Q5_K_M.bin -f your-audio.wav -l el --no-timestamps -np

Performance

Tested with Greek voice transcription:

✅ Excellent accuracy on Greek speech
✅ Fast inference on CPU (AMD GPU supported)
✅ Handles colloquial Greek well

Original Model

Original Safetensors version: https://huggingface.co/sam8000/whisper-large-v3-turbo-greek-greece

License

Same as original Whisper model (MIT)

Uploaded by: nicman23
Date: May 21, 2026

nicman23/whisper-large-v3-turbo-greek-gguf