training_args.json

{
    "adam_epsilon": 1e-08,
    "data_dir": "/content/data",
    "early_stop_callback": false,
    "eval_batch_size": 8,
    "fp_16": false,
    "gradient_accumulation_steps": 4,
    "learning_rate": 0.001,
    "max_grad_norm": 1.0,
    "max_input_length": 64,
    "model_name_or_path": "cl-tohoku/bert-base-japanese-whole-word-masking",
    "n_gpu": 1,
    "num_train_epochs": 2,
    "seed": 42,
    "shuffle_buffer_size": 65536,
    "tokenizer_name_or_path": "cl-tohoku/bert-base-japanese-whole-word-masking",
    "train_batch_size": 256,
    "warmup_ratio": 0.05,
    "weight_decay": 0.0
}