Sentiment Analysis on IMDB (F1-weighted)
89.8F1-weightedKEN
Evaluation results
| Method | Links | |
|---|---|---|
| KENBackbone=DeBERTa, Trainable params=107M, Reset params (%)=22.842024.02 | 89.8 | |
| KENBackbone=Ernie, Trainable params=80M, Reset params (%)=26.552024.02 | 89.6 | |
| KENBackbone=Ernie, Trainable params=69M, Reset params (%)=37.052024.02 | 89.4 | |
| KENBackbone=Ernie, Trainable params=75M, Reset params (%)=31.82024.02 | 89.2 | |
| KENBackbone=DeBERTa, Trainable params=99M, Reset params (%)=28.352024.02 | 88.3 | |
| KENBackbone=BLOOM1B7, Trainable params=664M, Reset params (%)=61.462024.02 | 87.6 | |
| KENBackbone=Bert, Trainable params=80M, Reset params (%)=26.552024.02 | 87.6 | |
| KENBackbone=Bert, Trainable params=75M, Reset params (%)=31.82024.02 | 87.3 | |
| KENBackbone=Bert, Trainable params=69M, Reset params (%)=37.052024.02 | 86.8 | |
| KENBackbone=BLOOM560M, Trainable params=429M, Reset params (%)=23.262024.02 | 85.8 | |
| KENBackbone=DistilBERT, Trainable params=51M, Reset params (%)=23.452024.02 | 84.6 | |
| KENBackbone=DistilBERT, Trainable params=47M, Reset params (%)=28.922024.02 | 84.4 | |
| KENBackbone=BLOOM560M, Trainable params=420M, Reset params (%)=24.82024.02 | 84.3 | |
| KENBackbone=Electra, Trainable params=14M, Reset params (%)=55.942024.02 | 84.3 | |
| KENBackbone=BLOOM1B7, Trainable params=531M, Reset params (%)=69.172024.02 | 84.2 | |
| KENBackbone=DistilBERT, Trainable params=44M, Reset params (%)=34.392024.02 | 83.2 | |
| KENBackbone=BLOOM560M, Trainable params=411M, Reset params (%)=26.342024.02 | 82.7 | |
| KENBackbone=DeBERTa, Trainable params=92M, Reset params (%)=33.862024.02 | 82.5 | |
| KENBackbone=Electra, Trainable params=12M, Reset params (%)=64.752024.02 | 82 | |
| KENBackbone=Electra, Trainable params=8.9M, Reset params (%)=75.562024.02 | 78.9 | |
| KENBackbone=BLOOM1B7, Trainable params=442M, Reset params (%)=74.312024.02 | 76.6 |