textattack/roberta-base-imdb

The model was fine-tuned for sequence classification using TextAttack and the imdb dataset loaded using the nlp library . The best score the model achieved on this task was 0.91436, as measured by the eval set accuracy, found after 2 epochs . For more information, check out TextAttack on Github and the TextAttack model Card on the model card on the current version of this article . The model has a learning rate of 3e-05 and a maximum sequence length of 128 . For example, the model was trained with a cross-entropy loss function and a batch size of 64 epochs with a learning rate of  3e-5 .,