Writing logs to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/xlnet-base-cased-rotten_tomatoes-2020-06-29-05:32/log.txt. Loading nlp dataset rotten_tomatoes, split train. Loading nlp dataset rotten_tomatoes, split validation. Loaded dataset. Found: 2 labels: ([0, 1]) Loading transformers AutoModelForSequenceClassification: xlnet-base-cased Tokenizing training data. (len: 8530) Tokenizing eval data (len: 1066) Loaded data and tokenized in 19.84232258796692s Training model across 4 GPUs ***** Running training ***** Num examples = 8530 Batch size = 16 Max sequence length = 128 Num steps = 2665 Num epochs = 5 Learning rate = 2e-05 Eval accuracy: 88.83677298311444% Best acc found. Saved model to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/xlnet-base-cased-rotten_tomatoes-2020-06-29-05:32/. Eval accuracy: 89.21200750469043% Best acc found. Saved model to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/xlnet-base-cased-rotten_tomatoes-2020-06-29-05:32/. Eval accuracy: 90.71294559099438% Best acc found. Saved model to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/xlnet-base-cased-rotten_tomatoes-2020-06-29-05:32/. Eval accuracy: 89.6810506566604% Eval accuracy: 89.6810506566604% Saved tokenizer to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/xlnet-base-cased-rotten_tomatoes-2020-06-29-05:32/. Wrote README to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/xlnet-base-cased-rotten_tomatoes-2020-06-29-05:32/README.md. Wrote training args to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/xlnet-base-cased-rotten_tomatoes-2020-06-29-05:32/train_args.json.