Azure Speech Studio finetune audio and text -- upload successfully with correct format but fails to train ("internal error" message and not known what errors)

Question

I follow the format to upload audio (in .wav, mono, 16khz) and text (in correct format stored in .txt).All the audios and text can upload to azure speech studio to fine tune, but when put into training phase, it fails. Training model as follows: What is the possible reasons for successful uploads but fails to train the model (audio+text)? What it means by the "internal error" error message during training?