Question Answering on CommonsenseQA (val)
82.8AccuracyLoRA
Evaluation results
| Method | Links | |
|---|---|---|
| LoRABackbone=Qwen2.5-7B-Instruct, Training Epochs=1, Scoring Method=CLL2026.02 | 82.8 | |
| D2-LoRABackbone=Qwen2.5-7B-Instruct, Training Epochs=1, Scoring Method=CLL2026.02 | 82.2 | |
| DoRABackbone=Qwen2.5-7B-Instruct, Training Epochs=1, Scoring Method=CLL2026.02 | 81.8 | |
| D2-LoRABackbone=Llama-3.2-3B-Instruct, Training Epochs=2, Scoring Method=CLL2026.02 | 75.4 | |
| DoRABackbone=Llama-3.2-3B-Instruct, Training Epochs=2, Scoring Method=CLL2026.02 | 74.8 | |
| LoRABackbone=Llama-3.2-3B-Instruct, Training Epochs=2, Scoring Method=CLL2026.02 | 73.4 | |
| AMOOptimizer=AMO, Model=Llama3.1-1.4B, Few-shot=02026.05 | 21.29 | |
| AMOOptimizer=AMO, Model=Llama3.1-760M, Few-shot=02026.05 | 21.13 |