Writing logs to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/bert-base-uncased-ag_news-2020-07-01-21:14/log.txt.
Loading nlp dataset ag_news, split train.
Loading nlp dataset ag_news, split test.
Loaded dataset. Found: 4 labels: ([0, 1, 2, 3])
Loading transformers AutoModelForSequenceClassification: bert-base-uncased
Tokenizing training data. (len: 120000)
Tokenizing eval data (len: 7600)
Loaded data and tokenized in 154.51052498817444s
Training model across 4 GPUs
***** Running training *****
	Num examples = 120000
	Batch size = 16
	Max sequence length = 128
	Num steps = 37500
	Num epochs = 5
	Learning rate = 3e-05
Eval accuracy: 94.22368421052632%
Best acc found. Saved model to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/bert-base-uncased-ag_news-2020-07-01-21:14/.
Eval accuracy: 94.5921052631579%
Best acc found. Saved model to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/bert-base-uncased-ag_news-2020-07-01-21:14/.
Eval accuracy: 94.85526315789473%
Best acc found. Saved model to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/bert-base-uncased-ag_news-2020-07-01-21:14/.
Eval accuracy: 95.14473684210526%
Best acc found. Saved model to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/bert-base-uncased-ag_news-2020-07-01-21:14/.
Eval accuracy: 94.64473684210526%
Saved tokenizer <textattack.models.tokenizers.auto_tokenizer.AutoTokenizer object at 0x7fc07b101eb0> to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/bert-base-uncased-ag_news-2020-07-01-21:14/.
Wrote README to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/bert-base-uncased-ag_news-2020-07-01-21:14/README.md.
Wrote training args to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/bert-base-uncased-ag_news-2020-07-01-21:14/train_args.json.