Tun-Wellens/whisper-medium-lb-included

Whisper Medium Luxembourgish Fine-Tuned Model

This is a fine-tuned version of OpenAI's Whisper medium model for Luxembourgish Automatic Speech Recognition (ASR).

Model Details

• Developed by: Tun Wellens
• Model type: Automatic Speech Recognition (ASR)
• Language(s): Luxembourgish (lb)
• License: apache-2.0
• Finetuned from model: openai/whisper-medium

Training Details

Fine-tuned using a combination of Luxembourgish datasets:

• Training datasets:
• https://huggingface.co/datasets/google/fleurs
• https://huggingface.co/datasets/mbarnig/lb-de-fr-en-pt-12800-TTS-CORPUS
• https://huggingface.co/datasets/Tun-Wellens/BSP-S4-lb_lu
• https://huggingface.co/datasets/Lemswasabi/luxembourgish-asr-rtl-lu
• Validation dataset: FLEURS (validation split)

Training was tracked using Weights & Biases (W&B) and performed with Hugging Face Transformers.

Intended Uses

• Primary use: Transcription of spoken Luxembourgish audio to text.

Limitations and Biases

• The model is specialized for Luxembourgish and does not not perform well on other languages.