SmallThinker-21BA3B-Instruct-GGUF

Home/Models/Mungert/SmallThinker-21BA3B-Instruct-GGUF

Mungert/SmallThinker-21BA3B-Instruct-GGUF

kind: gguf
quant: gguf (16-bit)
family: misc
modality: text-generation
root model: —
provider: Mungert
license: apache-2.0
language: en
context: 16K
library: transformers
vRAM (GB): —
size on disk: —
downloads: 321
likes: 3
created: —
last modified: 2025-09-24T15:44:47+00:00 (2025)
huggingface: https://huggingface.co/Mungert/SmallThinker-21BA3B-Instruct-GGUF
tags: transformersggufmoetext-generationenarxiv:2507.20984license:apache-2.0endpoints_compatibleregion:usconversational

benchmark	category	score	conf	evidence
IFEval	instruction-following	85.77	low	`\| SmallThinker-21BA3B-Instruct \| 84.43 \| <u>55.05</u> \| 82.4 \| 85.77 \| 60.3 \| <u>89.63</u> \| 76.26 \|`
MMLU	knowledge	84.43	low	`\| SmallThinker-21BA3B-Instruct \| 84.43 \| <u>55.05</u> \| 82.4 \| 85.77 \| 60.3 \| <u>89.63</u> \| 76.26 \|`
HumanEval	coding	82.93	low	`\| Gemma3-12b-it \| 78.52 \| 34.85 \| 82.4 \| 74.68 \| 44.5 \| 82.93 \| 66.31 \|`
MATH-500	math	82.4	low	`\| SmallThinker-21BA3B-Instruct \| 84.43 \| <u>55.05</u> \| 82.4 \| 85.77 \| 60.3 \| <u>89.63</u> \| 76.26 \|`
GPQA Diamond	reasoning	34.85	low	`\| Gemma3-12b-it \| 78.52 \| 34.85 \| 82.4 \| 74.68 \| 44.5 \| 82.93 \| 66.31 \|`

none scraped.