NRK-KIHUB/no-asr-eval-data-broadcast-sample

NRK Norwegian Speech Dataset (Sample) Dataset Description Note: This is a sample dataset containing a subset of chunks for demonstration and preview purposes. The full dataset is available privately. This dataset contains Norwegian speech data from NRK TV broadcasts (norge-rundt, supernytt), processed for automatic speech recognition (ASR) evaluation and research. Reference text caveat: Reference text is NRK's on-air teletext subtitling, not a verbatim… See the full description on the dataset page: https://huggingface.co/datasets/NRK-KIHUB/no-asr-eval-data-broadcast-sample.