Benchmark Almanac

MAmmoTH2-7B-Plus

MAmmoTH2-7B-Plus places in the top three on Ko-IFEval (Upstage).

10
Benchmarks with a score
0
Ranked first, out of five or more models
1
In the top three
TIGER-Lab (University of Waterloo) May 2024

MAmmoTH2-7B-Plus

Open weights Small

No Epoch Capabilities Index score yet
  • Ko-IFEval (Upstage) 44.6% #3 /21
  • Open Ko-LLM Leaderboard 2 38.3% #10 /21
  • Open Ko-LLM Leaderboard 42.8% #12 /18

How does it do against its rivals?

MAmmoTH2-7B-Plus beats Gemma 1.1 7B IT on 6 of the 8 benchmarks they share. Qwen2-7B does best against it, ahead on 7 of 9.

  1. Gemma 1.1 7B IT 6 ahead, 2 behind Compare
  2. Qwen1.5-14B 3 ahead, 6 behind Compare
  3. Qwen1.5-14B-Chat 3 ahead, 6 behind Compare
  4. Yi-1.5-9B-Chat 3 ahead, 6 behind Compare
  5. Qwen2-7B 2 ahead, 7 behind Compare

Sources: Official leaderboards, Epoch AI (CC BY 4.0) and labs’ model cards

Where it ranks

Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.

Knowledge and language

BenchmarkScoreRankSource
Ko-IFEval (Upstage) 44.6% #3 of 21 Official leaderboard
Open Ko-LLM Leaderboard 2 38.3% #10 of 21 Official leaderboard
Open Chinese LLM Leaderboard 53.6 #19 of 35 Official leaderboard
Open Portuguese LLM Leaderboard 64.1 #103 of 175 Official leaderboard
BLUEX 49% #115 of 177 Official leaderboard
Open Ko-LLM Leaderboard 42.8% #12 of 18 Official leaderboard
HeQ 52.7% #62 of 88 Official leaderboard
Open Hebrew LLM Leaderboard 38.3 #63 of 88 Official leaderboard
MMLU-Pro 40.9% #161 of 212 Official leaderboard

Other

BenchmarkScoreRankSource
Open LLM Leaderboard v1 67.8 #22 of 76 Official leaderboard

Who measured it?

Independent runs cover all 10 of its benchmarks.

  • An independent run only 10
Official leaderboards scored it on 10.

Sources: Official leaderboards, Epoch AI (CC BY 4.0) and labs’ model cards