Instruction Following on Vicuna-bench
8.24ScoreLlama-2-7B-star_instruct
Evaluation Results
| Method | Links | |
|---|---|---|
| Llama-2-7B-star_instructModel Scale=7B2024.11 | 8.24 | |
| Llama-2-7B-evol instructModel Scale=7B2024.11 | 7.27 | |
| JREGModel Size=3.4B, Alpha (α)=1.0, SFT=true2026.01 | 6.36 | |
| Llama-2-7B-alpacaModel Scale=7B2024.11 | 6.33 | |
| Pythia-1B-star_instructModel Scale=1B2024.11 | 5.93 | |
| BaselineModel Size=3.4B, SFT=true2026.01 | 5.89 | |
| Pythia-1B-RandomModel Scale=1B2024.11 | 5.13 | |
| Pythia-1B-evol_instructModel Scale=1B2024.11 | 5.07 | |
| Pythia-1B-IFDModel Scale=1B2024.11 | 4.6 | |
| Pythia-1B-alpacaModel Scale=1B2024.11 | 4.14 | |
| Sheared-LLaMA-1.3BModel Scale=1B2024.11 | 2.73 | |
| OPT-1.3BModel Scale=1B2024.11 | 2.49 | |
| Pythia-1BModel Scale=1B2024.11 | 1.68 |