LLaMA-13B
#250 of 274 scored.
MetaFeb 24, 2023Open weights
Capabilities Index
Range 89.0–106.4
100.6
Context
Tokens
—
Input
Per 1M tokens
—
Output
Per 1M tokens
—
Benchmarks
No independent scores yet.
Epoch AI hasn't run LLaMA-13B yet. Its scores show up here the morning after they do.
Everything else Epoch has for it
- TriviaQA77.9%
- LAMBADA75.2%
- HellaSwag72.3%
- PIQA60.2%
- Winogrande46.0%
- OpenBookQA41.9%
- ARC AI236.9%
- MMLU30.3%
- ScienceQA24.4%
- GSM8K20.5%
- BBH17.2%
On the index
Around it.
Sources