LLaMA-7B
#253 of 274 scored.
MetaFeb 24, 2023Open weights
Capabilities Index
Range 83.5–103.7
96.7
Context
Tokens
—
Input
Per 1M tokens
—
Output
Per 1M tokens
—
Benchmarks
No independent scores yet.
Epoch AI hasn't run LLaMA-7B yet. Its scores show up here the morning after they do.
Everything else Epoch has for it
- LAMBADA73.3%
- TriviaQA71.0%
- HellaSwag68.3%
- PIQA59.6%
- OpenBookQA42.9%
- Winogrande40.2%
- ARC AI230.1%
- ScienceQA14.9%
- MMLU14.1%
- BBH11.3%
- GSM8K11.0%
On the index
Around it.
Sources