Llama 2 7B — Benchmarks

Benchmark scores for Llama 2 7B aggregated from public leaderboards, with how it ranks among open models. See hardware requirements for what you need to run it.

Knowledge

BenchmarkScoreRank
MMLU45.3%#27 / 36
HellaSwag57.1%#16 / 22

Math

BenchmarkScoreRank
GSM8K14.6%#23 / 27

Reasoning

BenchmarkScoreRank
BIG-Bench Hard32.6%#9 / 11

Scores aggregated from public benchmark sources (each linked from the benchmark pages). llmrun does not run these benchmarks.