HF Model & Benchmark Matrix

24,078 variants · 14,984 scores · 120 benchmarks

Phi-3.5-mini-instruct

microsoft/Phi-3.5-mini-instruct

kind
original
quant
bf16 (16-bit)
family
phi
modality
text-generation
root model
microsoft/Phi-3.5-mini-instruct
provider
microsoft
license
mit
language
multilingual
context
,
library
transformers
vRAM (GB)
—
size on disk
—
downloads
955,420
likes
998
created
—
last modified
2025-12-10T20:22:56+00:00 (2025)
huggingface
https://huggingface.co/microsoft/Phi-3.5-mini-instruct
tags
transformerssafetensorsphi3text-generationnlpcodeconversationalcustom_codemultilingualarxiv:2404.14219arxiv:2407.13833arxiv:2403.06412license:miteval-resultstext-generation-inferenceendpoints_compatibleregion:usdeploy:sagemaker

Benchmark scores (5)

benchmarkcategoryscorecontextconfevidence
MATHmath86.2Mathmedium| Math | GSM8K (8-shot, CoT) | 86.2 | 54.4 | 84.2 | 82.4 | 84.9 | 82.4 | 91.3 |
TydiQAknowledge62.2MEGA TyDi QAmedium| MEGA TyDi QA | 62.2 | 55.9 | 54.0 | 63.7 | 54.5 | 65.6 | 63.6 | 81.8 |
MMLUknowledge55.4Multilingual MMLUmedium| Multilingual MMLU | 55.4 | 51.08 | 47.4 | 58.9 | 56.2 | 63.8 | 77.2 | 72.9 |
MGSMmath47.9MGSMmedium| MGSM | 47.9 | 41.56 | 31.8 | 63.3 | 56.7 | 75.1 | 75.8 | 81.7 |
MMLU-Proknowledge30.9Multilingual MMLU-Promedium| Multilingual MMLU-Pro | 30.9 | 30.21 | 15.0 | 34.0 | 21.4 | 43.0 | 57.9 | 53.2 |

Discussions (0)

none scraped.

Hugging Face model & benchmark data scraped & matrix-ified. 24,078 variants · 14,984 scores · 120 benchmarks.