Tiny QA Benchmark++: Ultra-Lightweight, Synthetic Multilingual Dataset Generation & Smoke-Tests for Continuous LLM Evaluation

Benchmark Model Rank Results
tinyqa-benchmark-on-tinyqabenchmark-core-engemma-3-4b#1Exact Match: 86.5
tinyqa-benchmark-on-tinyqabenchmark-core-enmistral-24b-instruct#2Exact Match: 84.6
tinyqa-benchmark-on-tinyqabenchmark-core-enllama-3.2-3b-instruct#3Exact Match: 84.6
tinyqa-benchmark-on-tinyqabenchmark-core-enministral-8b#4Exact Match: 80.8
tinyqa-benchmark-on-tinyqabenchmark-core-enministral-3b#5Exact Match: 76.9
tinyqa-benchmark-on-tinyqabenchmark-core-enllama-3.2-1b-instruct#6Exact Match: 53.8
tinyqa-benchmark-on-tinyqabenchmark-core-enmistral-7b-instruct#7Exact Match: 50.0
tinyqa-benchmark-on-tinyqabenchmark-core-engemma-3-12b#8Exact Macth: 90.4