KBLab/wav2vec2-large-voxrex-swedish

Automatic Speech Recognition·KBLab· 2.5M· 13

transformers cc0-1.0 315.5M params dataset:common_voicedataset:NST_Swedish_ASR_Databasedataset:P4

automatic speech recognition · transformers model

Open in MLForge Sign up free Desktop app Source ↗

# pull & run locally
pip install mlforge-sdk && mlforge pull KBLab/wav2vec2-large-voxrex-swedish

Model details

Task

Automatic Speech Recognition

Provider

KBLab

Framework

transformers

Parameters

315.5M

License

cc0-1.0

Downloads

2.5M

Likes

Paper

arXiv:2205.03026

Updated

2025-05-22

About KBLab/wav2vec2-large-voxrex-swedish

Finetuned version of KBs VoxRex large model using Swedish radio broadcasts, NST and Common Voice data. Evalutation without a language model gives the following: WER for NST + Common Voice test set (2% of total sentences) is 2.5%. WER for Common Voice test set is 8.49% directly and 7.37% with a 4-gram language model.

Related Automatic Speech Recognition

S pyannote/speaker-diarization-3.1 Automatic Speech Recognition 8.2M 2.5K 🤗 HF W argmaxinc/whisperkit-coreml Automatic Speech Recognition 8.1M 193 🤗 HF W openai/whisper-large-v3-turbo Automatic Speech Recognition ·808.9M params 7.1M 3.1K 🤗 HF W openai/whisper-base Automatic Speech Recognition ·72.6M params 6.2M 274 🤗 HF W jonatasgrosman/wav2vec2-large-xlsr-53-japanese Automatic Speech Recognition 6.1M 61 🤗 HF W openai/whisper-large-v3 Automatic Speech Recognition ·1.5B params 5.7M 5.9K 🤗 HF