MobileLLM-Pro

Home/Models/facebook/MobileLLM-Pro

facebook/MobileLLM-Pro

kind: original
quant: bf16 (16-bit)
family: misc
modality: text-generation
root model: facebook/MobileLLM-Pro-base
provider: facebook
license: fair-noncommercial-research-license
language: en
context: 128k
library: transformers
vRAM (GB): —
size on disk: —
downloads: 132
likes: 164
created: —
last modified: 2025-11-11T17:15:23+00:00 (2025)
huggingface: https://huggingface.co/facebook/MobileLLM-Pro
tags: transformerssafetensorsllama4_texttext-generationfacebookmetapytorchconversationalcustom_codeenarxiv:2511.06719base_model:facebook/MobileLLM-Pro-basebase_model:finetune:facebook/MobileLLM-Pro-baselicense:fair-noncommercial-research-licenseregion:us

benchmark	category	score	context	conf	evidence
PIQA	knowledge	76.55	PIQA	medium	`\| PIQA \| 76.55% \| 76.66% \| 75.52% \| 73.80% \| 75.14% \|`
BoolQ	knowledge	76.24	BoolQ	medium	`\| BoolQ \| 76.24% \| 77.49% \| 76.36% \| 63.20% \| 62.51% \|`
HellaSwag	knowledge	67.11	HellaSwag	medium	`\| HellaSwag \| 67.11% \| 64.89% \| 65.10% \| 62.30% \| 65.69% \|`
WinoGrande	knowledge	62.83	WinoGrande	medium	`\| WinoGrande \| 62.83% \| 62.43% \| 61.96% \| 58.20% \| 61.09% \|`
IFEval	instruction-following	62.0	IFEval	medium	`\| IFEval \| 62.0% \| 80.2% \| 59.5% \|`
HumanEval	coding	59.8	HumanEval	medium	`\| HumanEval \| 59.8% \| 41.5% \| 37.8% \|`
MBPP	coding	46.8	MBPP	medium	`\| MBPP \| 46.8% \| 35.2% \| 39.6% \|`
MMLU	knowledge	44.8	MMLU	medium	`\| MMLU \| 44.8% \| 29.9% \| 49.3% \|`
TriviaQA	knowledge	39.85	TriviaQA	medium	`\| TriviaQA \| 39.85% \| 37.26% \| 36.42% \| 39.80% \| 23.81% \|`
BFCL (function calling)	agentic	29.4	BFCL v2	medium	`\| BFCL v2 \| 29.4% \| \| 25.7% \|`

none scraped.