HF Model & Benchmark Matrix

24,078 variants · 14,984 scores · 120 benchmarks
Home/Orgs/meta-llama/meta-llama/Llama-3.1-8B

meta-llama/Llama-3.1-8B

family llama · 11 variants · unknown

provider
meta-llama
released
—
huggingface
https://huggingface.co/meta-llama/Llama-3.1-8B
best scores
MMLU: 84.4PopQA: 55.7TruthfulQA: 46.8BBH: 88.0DROP: 61.3MATH: 63.4GSM8K: 84.5HumanEval: 86.2IFEval: 72.8AlpacaEval: 12.4

Best scores per benchmark

benchmarkcategorybest score
BBHreasoning88.0
HumanEvalcoding86.2
GSM8Kmath84.5
MMLUknowledge84.4
IFEvalinstruction-following72.8
MATHmath63.4
DROPreasoning61.3
PopQAknowledge55.7
TruthfulQAsafety46.8
AlpacaEvalchat12.4

Variants (11)

variantkindquantbitsvRAMdlorgyear
Llama-3.1-8Boriginalbf1616.01,555,296meta-llama2024
Llama-3.1-Tulu-3-8B-SFToriginalbf1616.025,576allenai2025
Llama-3.1-8B-unsloth-bnb-4bitbnb-4bit4bit7,066unsloth2025
Llama-3.1-8B-bnb-4bitbnb-4bit4bit6,416unsloth2025
Llama-3.1-8Boriginalbf1616.02,994unsloth2025
UserLM-8boriginalbf1616.02,598microsoft2026
OpenMath2-Llama3.1-8Boriginalbf1616.01,383nvidia2024
Foundation-Sec-8B-GGUFggufgguf16.0280Mungert2025
Foundation-Sec-8B-Instruct-GGUFggufgguf16.0181Mungert2025
OpenMath2-Llama3.1-8B-GGUFggufgguf16.0147Mungert2025
UserLM-8b-GGUFggufgguf16.0147Mungert2025
Hugging Face model & benchmark data scraped & matrix-ified. 24,078 variants · 14,984 scores · 120 benchmarks.