eval: multi-model benchmark - nano is Pareto-optimal (0.896 micro F1 at 0.0116/doc)

bc61ea7

aditya0103 commited on 7 days ago