Experiments with wav2vec 2.0 models involving only 10 minutes of transcribed speech
Primary LanguageJupyter Notebook