Source code for the training pipeline of the text ranking model used in the paper entitled "Rank over Class: The Untapped Potential of Ranking in Natural Language Processing" (https://arxiv.org/abs/2009.05160).
First of all, thanks for sharing this great work!
It inspire me a lot ,cuz recent im meeting similar problem.
However, when im reading the code, I find there are no inference code.
So Im curious that how to covert the rank score to classification class. For example, I get 5 class for the sample.
But currently by the code, we can only get pair of rank scores for the pair of samples.