Sentiment Classification on SST-5 (32 samples)
50.2AccuracyPPT
Evaluation results
| Method | Links | |
|---|---|---|
| PPTModel Backbone=T5-XXL, Tuning Strategy=Prompt Tuning2021.09 | 50.2 | |
| Hybrid PPTModel Backbone=T5-XXL, Tuning Strategy=Prompt Tuning2021.09 | 50.1 | |
| Unified PPTModel Backbone=T5-XXL, Tuning Strategy=Prompt Tuning2021.09 | 46 | |
| T5-LargeTuning Strategy=Full-model Tuning2021.09 | 42.4 | |
| Hybrid PTModel Backbone=T5-XXL, Tuning Strategy=Prompt Tuning2021.09 | 40.9 | |
| T5-XXLTuning Strategy=Full-model Tuning2021.09 | 40.6 | |
| T5-XLTuning Strategy=Full-model Tuning2021.09 | 38.4 | |
| LM AdaptionModel Backbone=T5-XXL, Tuning Strategy=Prompt Tuning2021.09 | 36.2 | |
| Vanilla PTModel Backbone=T5-XXL, Tuning Strategy=Prompt Tuning2021.09 | 32.3 | |
| T5-SmallTuning Strategy=Full-model Tuning2021.09 | 31.1 | |
| T5-BaseTuning Strategy=Full-model Tuning2021.09 | 28.8 |