Rating Prediction on Amazon 5-core Office Products
0.5557MAEREVLoRA (Llama 1B)
Evaluation results
| Method | Links | ||
|---|---|---|---|
| REVLoRA (Llama 1B)Evaluation Protocol=Fine-Tuned, Optimization Loss=MAE2025.12 | 0.5557 | 0.8949 | |
| REVLoRA (Qwen 0.5B)Evaluation Protocol=Fine-Tuned, Optimization Loss=MAE2025.12 | 0.5572 | 0.8475 | |
| RGCLEvaluation Protocol=Deep Learning2025.12 | 0.6297 | 0.7381 | |
| DIRECTEvaluation Protocol=Deep Learning2025.12 | 0.6315 | 0.736 | |
| Llama 1BEvaluation Protocol=Zero-Shot2025.12 | 0.6412 | 1.1684 | |
| REVLoRA (Llama 1B MSE)Evaluation Protocol=Fine-Tuned, Optimization Loss=MSE2025.12 | 0.6454 | 0.7464 | |
| LURPEvaluation Protocol=Few-Shot2025.12 | 0.6455 | 1.2227 | |
| DeepCoNNEvaluation Protocol=Deep Learning2025.12 | 0.924 | 1.2262 |