Natural Language Explanation Generation on e-SNLI 60-shot
40.1AccuracySPARSEFIT
Evaluation results
| Method | Links | ||
|---|---|---|---|
| SPARSEFITFine-tuning Strategy=Att.Q+LN, Backbone=Llama 2-7B, Percentage Parameters=7.97%, Few-shot setting=60-shot2023.05 | 40.1 | 34.32 | |
| AdaLoRAFine-tuning Strategy=PEFT, Backbone=Llama 2-7B, Percentage Parameters=0.30%, Few-shot setting=60-shot2023.05 | 38.15 | 33.48 | |
| Full Fine-tuningFine-tuning Strategy=Baseline, Backbone=Llama 2-7B, Percentage Parameters=100%, Few-shot setting=60-shot2023.05 | 34.91 | 28.62 |