elgeish/wav2vec2-large-lv60-timit-asr

The model can be used directly (without a language model) as follows: import load_dataset, transformers import Wav2Vec2ForCTC and processor . Make sure that your speech input is sampled at 16kHz on the timit_asr dataset . The model is fine-tuned using the soundfile and torch . The output is based on the sound file and the language model of the model . The soundfile can be downloaded from soundfile as sf and sf to test the model's accuracy . For example, the model is used to produce the output of a fine tuning version of this version of the fine tuning script . It is available in the,