Phi-mini-MoE-instruct

Home/Models/microsoft/Phi-mini-MoE-instruct

microsoft/Phi-mini-MoE-instruct

kind: original
quant: bf16 (16-bit)
family: misc
modality: text-generation
root model: microsoft/Phi-mini-MoE-instruct
provider: microsoft
license: mit
language: en
context: 4k
library: transformers
vRAM (GB): —
size on disk: —
downloads: 179,446
likes: 38
created: —
last modified: 2025-12-10T18:20:28+00:00 (2025)
huggingface: https://huggingface.co/microsoft/Phi-mini-MoE-instruct
tags: transformerssafetensorsphimoetext-generationconversationalcustom_codeenarxiv:2506.18349arxiv:2404.14219arxiv:2409.12136license:mitregion:us

benchmark	category	score	conf	evidence
GSM8K	math	84.89	high	`\| Phi-mini-MoE \| 7.6B \| 2.4B \| 70.68 \| 49.68 \| 55.27 \| 84.91 \| 73.80 \| 84.89 \| 7.59 \|`
MMLU	knowledge	70.68	high	`\| Phi-mini-MoE \| 7.6B \| 2.4B \| 70.68 \| 49.68 \| 55.27 \| 84.91 \| 73.80 \| 84.89 \| 7.59 \|`
BBH	reasoning	55.27	high	`\| Phi-mini-MoE \| 7.6B \| 2.4B \| 70.68 \| 49.68 \| 55.27 \| 84.91 \| 73.80 \| 84.89 \| 7.59 \|`
MMLU-Pro	knowledge	49.68	high	`\| Phi-mini-MoE \| 7.6B \| 2.4B \| 70.68 \| 49.68 \| 55.27 \| 84.91 \| 73.80 \| 84.89 \| 7.59 \|`
MT-Bench	chat	7.59	high	`\| Phi-mini-MoE \| 7.6B \| 2.4B \| 70.68 \| 49.68 \| 55.27 \| 84.91 \| 73.80 \| 84.89 \| 7.59 \|`

none scraped.