Are Transformers in Pre-trained LM A Good ASR Encoder? An Empirical Study

Re-assign community