HF Model & Benchmark Matrix

24,078 variants · 14,984 scores · 120 benchmarks
Home/Orgs/inclusionAI/inclusionAI/Ling-lite

inclusionAI/Ling-lite

family misc · 7 variants · unknown

provider
inclusionAI
released
—
huggingface
https://huggingface.co/inclusionAI/Ling-lite
best scores
MMLU: 74.87GPQA: 40.91HumanEval: 89.02LiveCodeBench: 31.9MATH: 79.12AIME 2024: 56.3Math Olympiad: 37.33BBH: 74.58IFEval: 81.09AlpacaEval: 3.69OpenBookQA: 61.98WER LibriSpeech: 1.44WER VoxPopuli: 5.8MMBench: 80.8MMMU: 56.3MathVista: 68.1ChartQA: 85.1DocVQA: 93.0MATH-500: 93.8GPQA Diamond: 46.5

Best scores per benchmark

benchmarkcategorybest score
MATH-500math93.8
DocVQAmultimodal93.0
HumanEvalcoding89.02
ChartQAmultimodal85.1
IFEvalinstruction-following81.09
MMBenchmultimodal80.8
MATHmath79.12
MMLUknowledge74.87
BBHreasoning74.58
MathVistamultimodal68.1
OpenBookQAknowledge61.98
AIME 2024math56.3
MMMUmultimodal56.3
GPQA Diamondreasoning46.5
GPQAreasoning40.91
Math Olympiadmath37.33
LiveCodeBenchcoding31.9
WER VoxPopulispeech-asr5.8
AlpacaEvalchat3.69
WER LibriSpeechspeech-asr1.44

Variants (7)

variantkindquantbitsvRAMdlorgyear
Ling-liteoriginalbf1616.01,085inclusionAI2025
inclusionAI_Ling-lite-0415-GGUFggufgguf2.0907bartowski2025
Ling-lite-0415-GGUFggufgguf3.0677lmstudio-community2025
inclusionAI_Ling-lite-GGUFggufgguf2.0566bartowski2025
Ming-Lite-Omnioriginalbf1616.082inclusionAI2025
Ring-lite-distill-previeworiginalbf1616.033inclusionAI2025
Ring-lite-linear-previeworiginalbf1616.024inclusionAI2025
Hugging Face model & benchmark data scraped & matrix-ified. 24,078 variants · 14,984 scores · 120 benchmarks.