Mathematical Reasoning on GSM8K (test) (loss)
0.491LossLORA-GA
Evaluation results
| Method | Links | |
|---|---|---|
| LORA-GABackbone=Llama-3.2-3B, Evaluation Protocol=fine-tuning, Epoch=22026.05 | 0.491 | |
| LoRABackbone=Llama-3.2-3B, Evaluation Protocol=fine-tuning, Epoch=22026.05 | 0.492 | |
| DORABackbone=Llama-3.2-3B, Evaluation Protocol=fine-tuning, Epoch=22026.05 | 0.492 | |
| BALORABackbone=Llama-3.2-3B, Evaluation Protocol=fine-tuning, Epoch=22026.05 | 0.493 | |
| DORABackbone=Llama-3.2-3B, Evaluation Protocol=fine-tuning, Epoch=12026.05 | 0.497 | |
| LoRABackbone=Llama-3.2-3B, Evaluation Protocol=fine-tuning, Epoch=12026.05 | 0.498 | |
| OLORABackbone=Llama-3.2-3B, Evaluation Protocol=fine-tuning, Epoch=22026.05 | 0.503 | |
| LORA-GABackbone=Llama-3.2-3B, Evaluation Protocol=fine-tuning, Epoch=12026.05 | 0.504 | |
| BALORABackbone=Llama-3.2-3B, Evaluation Protocol=fine-tuning, Epoch=12026.05 | 0.506 | |
| OLORABackbone=Llama-3.2-3B, Evaluation Protocol=fine-tuning, Epoch=12026.05 | 0.51 |