GetmanY1/wav2vec2-xls-r-300m-sami-parl-ext-ft
Automatic Speech Recognition
•
Updated
•
8
wav2vec 2.0 Northern Sámi ASR models fine-tuned on Sámi Parliament speech (https://www.isca-archive.org/interspeech_2024/getman24b_interspeech.pdf)
Note This is the best-performing model in this collection. However, for improved performance, consider switching to models from https://huggingface.co/collections/GetmanY1/wav2vec2-sami-22k-66ead12fe465d6302b63d11b , which have been pre-trained on 22.4K hours of Sami speech.