Benchmark Almanac

Llama 3.2 3B

Llama 3.2 3B.

Meta Released 24 Sep 2024
25
Benchmarks with a score
0
Ranked first, out of five or more models
0
In the top three
Meta Sep 2024

Llama 3.2 3B

Open weights

No Epoch Capabilities Index score yet
  • Vectara Hallucination Leaderboard (HHEM) 7.90% #32 /148
  • BigCodeBench-Hard 5.40% #92 /96
  • AlGhafa 42.9% #31 /39

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
Vectara Hallucination Leaderboard (HHEM) 7.90% #32 of 148 Official leaderboard
MadinahQA 38.7% #25 of 39 Official leaderboard
EQ-Bench v2 61.1 #65 of 85 Official leaderboard
ArabicMMLU 35.7% #30 of 39 Official leaderboard
Open Arabic LLM Leaderboard v2 41.1% #30 of 39 Official leaderboard
AlGhafa 42.9% #31 of 39 Official leaderboard
ArbMMLU-HT 26% #31 of 39 Official leaderboard
ALRAGE 53.3% #33 of 39 Official leaderboard
FaithJudge 46.2% #33 of 37 Official leaderboard
LMArena (Elo) 1,167 #299 of 333 Official leaderboard
MMLU-Pro 22.2% #194 of 212 Official leaderboard
BuzzBench 26.6 #22 of 23 Official leaderboard
Creative Writing v3 42.8 #115 of 120 Official leaderboard
WritingBench 38.1 #44 of 45 Official leaderboard
Wer wird Millionär? LLM Benchmark 50.0 #76 of 76 Official leaderboard
SEA-HELM 25.1 #62 of 62 Official leaderboard

Math and science

BenchmarkScoreRankSource
ZebraLogic 7.40% #42 of 46 Official leaderboard

Coding

BenchmarkScoreRankSource
BigCodeBench 25.9% #63 of 69 Official leaderboard
BigCodeBench-Hard 5.40% #92 of 96 Official leaderboard

Agents and tools

BenchmarkScoreRankSource
Berkeley Function Calling Leaderboard (BFCL) 45.9% #56 of 67 Official leaderboard
BFCL V4 21.9% #57 of 65 Official leaderboard
BALROG 10.1% #32 of 35 Official leaderboard
LisanBench 8.00 #132 of 133 Official leaderboard
AceBench 19.6% #16 of 16 Official leaderboard

Safety and security

BenchmarkScoreRankSource
AraTrust 58.5% #27 of 39 Official leaderboard