New dataset

[iuria21]

Hi, thanks for your work. This is a question, not an issue, so feel free to close it if you want.

I have a dataset with not-labelled call transcriptions, and I want to train a classifier for them. I'm wondering if I could use the adversarial training to train it (once part of the dataset it's labelled manually).

I'm waiting your suggestion, thanks again!