from speechbrain.pretrained import EncoderDecoderASR
asr_model = EncoderDecoderASR.from_hparams(source="speechbrain/asr-crdnn-rnnlm-librispeech", savedir="pretrained_models/asr-crdnn-rnnlm-librispeech")
asr_model.transcribe_file('speechbrain/asr-crdnn-rnnlm-librispeech/example.wav')