Question Answering on TextbookQA (dev)
52.5F1 ScoreFewshotQA w/ MINPROMPT
Evaluation results
| Method | Links | |
|---|---|---|
| FewshotQA w/ MINPROMPTNumber of training examples=128, Number of Parameters=406M2023.10 | 52.5 | |
| FewshotQA w/ MINPROMPTNumber of training examples=64, Number of Parameters=406M2023.10 | 48.7 | |
| FewshotQA w/ MINPROMPTNumber of training examples=32, Number of Parameters=406M2023.10 | 46.5 | |
| FewshotQANumber of training examples=128, Number of Parameters=406M2023.10 | 46.2 | |
| PMRNumber of training examples=128, Number of Parameters=406M2023.10 | 45.1 | |
| FewshotQANumber of training examples=64, Number of Parameters=406M2023.10 | 44.8 | |
| Splinter w/ MINPROMPTNumber of training examples=128, Number of Parameters=110M2023.10 | 44.2 | |
| Splinter w/ MINPROMPTNumber of training examples=64, Number of Parameters=110M2023.10 | 42.6 | |
| SplinterNumber of training examples=128, Number of Parameters=110M2023.10 | 42.6 | |
| FewshotQA w/ MINPROMPTNumber of training examples=16, Number of Parameters=406M2023.10 | 42.2 | |
| PMRNumber of training examples=64, Number of Parameters=406M2023.10 | 41.8 | |
| FewshotQANumber of training examples=32, Number of Parameters=406M2023.10 | 41.7 | |
| Splinter w/ MINPROMPTNumber of training examples=32, Number of Parameters=110M2023.10 | 38.2 | |
| Splinter w/ MINPROMPTNumber of training examples=16, Number of Parameters=110M2023.10 | 37 | |
| PMRNumber of training examples=32, Number of Parameters=406M2023.10 | 36.4 | |
| SplinterNumber of training examples=64, Number of Parameters=110M2023.10 | 35.9 | |
| FewshotQANumber of training examples=16, Number of Parameters=406M2023.10 | 33.1 | |
| PMRNumber of training examples=16, Number of Parameters=406M2023.10 | 31 | |
| SplinterNumber of training examples=32, Number of Parameters=110M2023.10 | 27.6 | |
| SpanBERTNumber of training examples=128, Number of Parameters=110M2023.10 | 20.9 | |
| SplinterNumber of training examples=16, Number of Parameters=110M2023.10 | 19.4 | |
| SpanBERTNumber of training examples=64, Number of Parameters=110M2023.10 | 13 | |
| RoBERTaNumber of training examples=128, Number of Parameters=110M2023.10 | 8.2 | |
| SpanBERTNumber of training examples=32, Number of Parameters=110M2023.10 | 7.6 | |
| SpanBERTNumber of training examples=16, Number of Parameters=110M2023.10 | 7.5 | |
| RoBERTaNumber of training examples=64, Number of Parameters=110M2023.10 | 5.4 | |
| RoBERTaNumber of training examples=32, Number of Parameters=110M2023.10 | 4.3 | |
| RoBERTaNumber of training examples=16, Number of Parameters=110M2023.10 | 3.3 |