X

XLSR 300M Bokmaal

Developed by NbAiLab
This model is a speech recognition model fine-tuned on Norwegian Bokmål (NPSC dataset) based on facebook/wav2vec2-xls-r-300m, achieving a test set WER of 7.7%
Downloads 26
Release Time : 3/2/2022

Model Overview

An automatic speech recognition (ASR) model optimized for Norwegian Bokmål, fine-tuned based on the XLSR-300M architecture

Model Features

Low word error rate
Achieves 7.7% WER and 2.8% CER on Norwegian Bokmål test set
Domain-specific optimization
Trained on the Norwegian Parliamentary Speech Corpus (NPSC), suitable for formal speech recognition
Efficient fine-tuning
Obtained through 15 rounds of fine-tuning on the base XLSR-300M model

Model Capabilities

Norwegian Bokmål speech-to-text
16kHz audio processing
Long-form speech recognition

Use Cases

Government/Parliament
Parliament meeting minutes
Automatically transcribe Norwegian parliamentary meeting content
92.3% accuracy (WER 7.7%)
Education
Lecture recording transcription
Convert Norwegian educational audio into text transcripts
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
© 2025AIbase