openai/whisper-medium

This model is a fine-tuned version of openai/whisper-medium on the samromur, malromur, raddromur and althingi datasets. It achieves the following results on the evaluation set, the output is lowercased and punctuation is removed:

Google Fleurs 13.94% WER
Samrómur 10.08% WER

Framework versions

Transformers 4.26.0.dev0
Pytorch 1.13.0+cu117
Datasets 2.7.1.dev0
Tokenizers 0.13.2

Downloads last month: 6

Inference API

Automatic Speech Recognition

This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Datasets used to train DavidErikMollberg/whipser-medium-is

Evaluation results

WER on samromur
test set self-reported

10.08%
WER on google/fleurs
test set self-reported

13.940

View on Papers With Code