General Evaluation on Instruction Tuning Suite (BIG-bench Hard, MMLU, TyDi QA, MGSM)
74.1Average ScoreFlan-PaLM 2 (L)
Evaluation results
| Method | Links | |
|---|---|---|
| Flan-PaLM 2 (L)2023.05 | 74.1 | |
| PaLM 2 (L)2023.05 | 69.3 | |
| Flan-U-PaLM-540B2023.05 | 66.1 | |
| U-PaLM-540B2023.05 | 58.6 |