HF Model & Benchmark Matrix

24,078 variants · 14,984 scores · 120 benchmarks
Home/Orgs/NousResearch/NousResearch/Meta-Llama-3-8B-Instruct

NousResearch/Meta-Llama-3-8B-Instruct

family llama · 3 variants · unknown

provider
NousResearch
released
—
huggingface
https://huggingface.co/NousResearch/Meta-Llama-3-8B-Instruct
best scores
IFEval: 44.69BBH: 28.48MATH: 8.31GPQA: 5.59MUSR: 8.38MMLU-Pro: 29.65

Best scores per benchmark

benchmarkcategorybest score
IFEvalinstruction-following44.69
MMLU-Proknowledge29.65
BBHreasoning28.48
MUSRreasoning8.38
MATHmath8.31
GPQAreasoning5.59

Variants (3)

variantkindquantbitsvRAMdlorgyear
ChimeraLlama-3-8B-v3mergemerge8,029mlabonne2024
ChimeraLlama-3-8B-v2mergemerge8,021mlabonne2024
Llama-3-8B-Instruct-Coder-v2-GGUFggufgguf1.0950bartowski2024
Hugging Face model & benchmark data scraped & matrix-ified. 24,078 variants · 14,984 scores · 120 benchmarks.