Skip to content

Benchmark

Persian-family NLI Macro-F1 (Iranian Persian aggregate, zero-shot)

Multilingual benchmarks compare model performance across languages. Higher scores are better.

Models measured
2
Scores recorded
2
Sources
1
IQ weight
No weight published

Score distribution

55.2
Mean
59.3
Median
4.2
Std dev
51
10th percentile
59.3
90th percentile
51
Min
5159.3
The band covers the middle 80% of scores. The line marks the median.

Leaderboard

#ModelScore
1Qwen2.5 3B Instructtable: 2, task: NLI, metric: Macro-F1, subset: Iranian Persian aggregate· Persian, Dari and Tajiki paper data docs59.3
2Qwen2.5 1.5B Instructtable: 2, task: NLI, metric: Macro-F1, subset: Iranian Persian aggregate· Persian, Dari and Tajiki paper data docs51

Data credits: Noor Mairukh Khan Arnob and Abu Bakar Siddique Mahi. One Language, Three of Its Voices, SilkRoadNLP 2026, ©2026 ACL, CC BY 4.0. Numeric/protocol extract from Tables 2, 5 and 6; language and pooled subsets retained. (licence)