chinese summarization

[Tangjiandd]

I use chinese text summarization datasets to train this model.
but when I test the trained model, the result will be ", , , ....,"

[ByUnal]

Yes, I got the similar results when I tried this model for Turkish. I think the model doesn't work for other languages, or we're doing a mistake in tokenization part