in the loop

Stable Beluga 2
#210 of 274 scored.

Stability AIJul 20, 2023Open weights

Capabilities Index

Range 108.4–120.1

117.1

Context

Tokens

—

Input

Per 1M tokens

—

Output

Per 1M tokens

—

Benchmarks

No independent scores yet.

Epoch AI hasn't run Stable Beluga 2 yet. Its scores show up here the morning after they do.

Everything else Epoch has for it

  • ARC AI281.5%
  • HellaSwag78.8%
  • LAMBADA71.3%
  • GSM8K69.6%
  • PIQA66.6%
  • BBH59.1%
  • MMLU58.1%

On the index

Around it.

  1. #207Ministral 3BMistral118.1
  2. #208Yi-34B01.AI117.4
  3. #209phi-3-mini 3.8BMicrosoft117.4
  4. #210Stable Beluga 2Stability AI117.1
  5. #211Gemini 1.0 ProGoogle117.0
  6. #212Llama 3.1-8BMeta116.6
  7. #213Llama 3-8BMeta116.5

Sources