the model cannot separate the mix speech!

[JayYang-Fdu]

Hello, I am disturbing you. I used my timit dataset to find that the trained model could not separate mixed speech. May I venture to ask, do you have a trained model for reference? Thank you very much!