Language Modeling on Wikitext-2 raw v1
2.241LossOLORA
Evaluation results
| Method | Links | |
|---|---|---|
| OLORABackbone=Llama-3.2-3B, Fine-tuning mode=fine-tuning2026.05 | 2.241 | |
| BALORABackbone=Qwen-2.5-3B, Fine-tuning mode=fine-tuning2026.05 | 2.251 | |
| LORABackbone=Qwen-2.5-3B, Fine-tuning mode=fine-tuning2026.05 | 2.257 | |
| DORABackbone=Qwen-2.5-3B, Fine-tuning mode=fine-tuning2026.05 | 2.257 | |
| LORA-GABackbone=Qwen-2.5-3B, Fine-tuning mode=fine-tuning2026.05 | 2.264 | |
| BALORABackbone=Llama-3.2-3B, Fine-tuning mode=fine-tuning2026.05 | 2.274 | |
| OLORABackbone=Qwen-2.5-3B, Fine-tuning mode=fine-tuning2026.05 | 2.276 | |
| LORA-GABackbone=Llama-3.2-3B, Fine-tuning mode=fine-tuning2026.05 | 2.276 | |
| LORABackbone=Llama-3.2-3B, Fine-tuning mode=fine-tuning2026.05 | 2.278 | |
| DORABackbone=Llama-3.2-3B, Fine-tuning mode=fine-tuning2026.05 | 2.278 |