Commonsense Reasoning on BigBenchHard
71.7AccuracyPhi-3 Mini-4k
Evaluation results
| Method | Links | |
|---|---|---|
| Phi-3 Mini-4kShots=0-shot2024.10 | 71.7 | |
| Mistral 8x7BParameters=8x7B, Shots=0-shot2024.10 | 69.7 | |
| Gemma 7BParameters=7B, Shots=0-shot2024.10 | 59.6 | |
| Shakti-LLMParameters=2.5B, Shots=0-shot2024.10 | 58.2 | |
| Mistral 7BParameters=7B, Shots=0-shot2024.10 | 57.3 | |
| Llama 3 8BParameters=8B, Shots=0-shot2024.10 | 51.5 | |
| T-SPINIteration=Iter32026.01 | 45.27 | |
| T-SPINIteration=Iter42026.01 | 45 | |
| SFT2026.01 | 44.9 | |
| T-SPINIteration=Iter12026.01 | 44.82 | |
| Zephyr-7B2026.01 | 44.6 | |
| SPINIteration=Iter12026.01 | 44.59 | |
| T-SPINIteration=Iter22026.01 | 44.48 | |
| SPINIteration=Iter02026.01 | 44.41 | |
| T-SPINIteration=Iter02026.01 | 44.17 | |
| SPINIteration=Iter22026.01 | 43.95 | |
| SPINIteration=Iter32026.01 | 43.19 | |
| SPINIteration=Iter42026.01 | 43.1 |