Chinese Language Understanding on MMMLU
37.08MMMLU ScoreIRR
Evaluation Results
| Method | Links | |
|---|---|---|
| IRRfine-tuning dataset=Alpaca Chinese2024.12 | 37.08 | |
| SFTfine-tuning dataset=Alpaca Chinese2024.12 | 36.85 | |
| IRRdfine-tuning dataset=Alpaca Chinese2024.12 | 36.82 | |
| DAREfine-tuning dataset=Alpaca Chinese2024.12 | 36.78 | |
| RESTAfine-tuning dataset=Alpaca Chinese2024.12 | 33.03 | |
| RESTAdfine-tuning dataset=Alpaca Chinese2024.12 | 32.4 | |
| SafeDecodingfine-tuning dataset=Alpaca Chinese2024.12 | 23.29 | |
| Safe LoRAfine-tuning dataset=Alpaca Chinese2024.12 | 22.61 |