espnet2.sds.end_to_end.mini_omni.inference.A1_T1

source

espnet2.sds.end_to_end.mini_omni.inference.A1_T1(fabric, audio_feature, input_ids, leng, model, text_tokenizer, step)