StableBeluga2 — Benchmarks

Benchmark scores for StableBeluga2 aggregated from public leaderboards, with how it ranks among open models. See hardware requirements for what you need to run it.

Overall rank: #4 of 75 open modelscomposite 69.1/100 across 4 benchmarks in 3 categories · methodology

Knowledge

BenchmarkScoreOpen rankAll models
MMLU68.6%#30 / 76#71 / 136
HellaSwag84.1%#9 / 42#14 / 76

Math

BenchmarkScoreOpen rankAll models
GSM8K69.6%#18 / 59#27 / 93

Reasoning

BenchmarkScoreOpen rankAll models
BIG-Bench Hard69.3%#8 / 37#12 / 50

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