textattack/distilbert-base-uncased-RTE

The best score the model achieved on this task was 0.6570397111913358, as measured by the eval set accuracy, found after 4 epochs . The model was fine-tuned for 5 epochs with a batch size of 16, a learning rate of 2e-05, and a maximum sequence length of 128 . For more information, check out TextAttack on Github and the nlp library on the current version of TextAttack . For example, the model was trained with a cross-entropy loss function and the glue dataset was loaded using TextAttack and the library's nlp dataset was used to train the model . It is now available on the,