Cseti/Qwen3-ASR_Hungarian_v1

Qwen3-ASR Hungarian Full Fine-tune

This repository contains a fully fine-tuned version of the Qwen3-ASR model.

Training

This model was trained on RunPod cloud GPUs.

Performance Comparison

Using 1000 samples from CommonVoice 17 as the evaluation dataset, the following metrics demonstrate a significant improvement over the base model:

Metric	Base Model (Original)	Fine-tuned Model
Raw WER	29.20%	16.83%
Normalized WER	26.24%	13.68%

Non-Commercial Use Only

Due to the specific licensing and characteristics of the dataset used during the fine-tuning process, this model is strictly prohibited for commercial use. It is intended for research, evaluation.

Support

Producing and sharing this kind of open-source work requires renting cloud GPUs, which gets expensive quickly. If you find it useful and would like me to keep contributing, your support is very much appreciated: