Flex-math-2x7B-1T

Home/Models/allenai/Flex-math-2x7B-1T

allenai/Flex-math-2x7B-1T

kind: original
quant: bf16 (16-bit)
family: misc
modality: text-generation
root model: allenai/Flex-math-2x7B-1T
provider: allenai
license: apache-2.0
language: en
context: —
library: transformers
vRAM (GB): —
size on disk: —
downloads: 135
likes: 4
created: —
last modified: 2026-03-02T19:12:45+00:00 (2026)
huggingface: https://huggingface.co/allenai/Flex-math-2x7B-1T
tags: transformerssafetensorsflex_olmotext-generationmoeolmoflexolmoconversationalenarxiv:2507.00000license:apache-2.0co2_eq_emissionsendpoints_compatibleregion:us

benchmark	category	score	context	conf	evidence
MATH	math	62.5	Math	medium	`\| Math \| 62.5 \| 44.3 \| 50.6 \| 24.1 \| 42.0 \| 45.6 \| 53.1 \| 42.6 \| 28.0 \| 50.7 \| 15.8 \| 41.8 \|`
MMLU	knowledge	50.6		high	`\| Math \| 62.5 \| 44.3 \| 50.6 \| 24.1 \| 42.0 \| 45.6 \| 53.1 \| 42.6 \| 28.0 \| 50.7 \| 15.8 \| 41.8 \|`
BBH	reasoning	45.6		high	`\| Math \| 62.5 \| 44.3 \| 50.6 \| 24.1 \| 42.0 \| 45.6 \| 53.1 \| 42.6 \| 28.0 \| 50.7 \| 15.8 \| 41.8 \|`
AGIEval	reasoning	42.0		high	`\| Math \| 62.5 \| 44.3 \| 50.6 \| 24.1 \| 42.0 \| 45.6 \| 53.1 \| 42.6 \| 28.0 \| 50.7 \| 15.8 \| 41.8 \|`
MMLU-Pro	knowledge	24.1		high	`\| Math \| 62.5 \| 44.3 \| 50.6 \| 24.1 \| 42.0 \| 45.6 \| 53.1 \| 42.6 \| 28.0 \| 50.7 \| 15.8 \| 41.8 \|`

none scraped.