Gyuwan Kim

Results 2 comments of Gyuwan Kim

Agreed. It could be okay for the benchmark dataset but seems problematic in a real scenario.

You can calcuate log probability of given sentence using trained model.