valhalla/s2t_librispeech_small

The following script shows how to evaluate this model on the LibriSpeech "clean" and "other" test dataset . The script uses the model to evaluate the model on a soundfile and other test datasets . The model is called Speech2TextTransformerForConditionalGeneration and tokenizer . The code is based on the soundfile as sf and the sound file as soundfile . The test dataset is "clean," "other", "clean", "test", and "test" for other test dataset and "other test takes it to evaluation on Librispeech_eval. The script is now available to download and,