pariajm/deep-disfluency-detector

Experimental setup

sl403 opened this issue · 3 comments

sl403 commented

Hi,

What's your experimental setup?
Is python2? What's your tensorflow version?
thank you.

Hi,

What's your experimental setup?
Is python2? What's your tensorflow version?
thank you.

Hi,
It should work with python 3 and tensorflow > 0.12.
I updated the repo (fixed print commands).

sl403 commented

Thank for your reply.

Another question unrelated to the title.
Can you release some trained model and the predict code make me to predict it?
If possible, can you provide your experimental test data?
Thank you.

Thank for your reply.

Another question unrelated to the title.
Can you release some trained model and the predict code make me to predict it?
If possible, can you provide your experimental test data?
Thank you.

Since Switchboard is not open-source, I cannot release the test set (you can follow the pre-processing steps and use the exact test set split explained in the paper though).
But I just released the trained model and the prediction code. Check the repo, please!