Generative Classification on DBpedia (test)
96.9AccuracyPrefixMemory-Tuning
Evaluation results
| Method | Links | |
|---|---|---|
| PrefixMemory-TuningBackbone Model=Qwen2.5-3B-Instruct2025.06 | 96.9 | |
| Full fine-tuningBackbone Model=Qwen2.5-3B-Instruct2025.06 | 94.4 | |
| LoRA+Backbone Model=Qwen2.5-3B-Instruct2025.06 | 94.4 | |
| PrefixMemory-TuningBackbone Model=LLaMA2-7B-Chat2025.06 | 92.7 | |
| Full fine-tuningBackbone Model=LLaMA2-7B-Chat2025.06 | 92.6 | |
| LoRA+Backbone Model=LLaMA2-7B-Chat2025.06 | 91.9 | |
| LoRABackbone Model=LLaMA2-7B-Chat, rank=r=642025.06 | 90.1 | |
| LoRABackbone Model=Qwen2.5-3B-Instruct, rank=r=642025.06 | 89.5 | |
| Prefix-TuningBackbone Model=Qwen2.5-3B-Instruct, virtual tokens=m=322025.06 | 82 | |
| Prefix-TuningBackbone Model=LLaMA2-7B-Chat, virtual tokens=m=322025.06 | 61.3 |