mrm8488/t5-base-finetuned-quarel

The model was presented in Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer . Google's T5 fine-tuned on QuaRel for QA downstream task . The training script is a slightly modified version of this awesome one by Suraj Patil . The dataset has 2771 questions relating 19 different types of quantities . For example, "Jenny observes that the robot vacuum cleaner moves slower on the living room carpet than on the bedroom carpet. Which carpet has more friction?" The first of these models (called QuaSP+) significantly outperforms off-the-shelf tools . The second (QuaSP+Zero) demonstrates zero-shot,