--- datasets: - stanfordnlp/sst2 language: - en metrics: - accuracy: 0.91789 ---