./run_mlm_flax.py \
    --output_dir="./" \
    --model_type="roberta" \
    --config_name="./" \
    --tokenizer_name="./" \
    --max_seq_length="128" \
    --weight_decay="0.01" \
    --train_file="/mnt/disks/flaxdisk/divided_corpus/train-shard-0001-of-0001.json" \
    --validation_file="/mnt/disks/flaxdisk/divided_corpus/validation-shard-0001-of-0001.json" \
    --per_device_train_batch_size="128" \
    --per_device_eval_batch_size="128" \
    --learning_rate="3e-4" \
    --warmup_steps="5000" \
    --overwrite_output_dir \
    --num_train_epochs="29" \
    --adam_beta1="0.9" \
    --adam_beta2="0.98" \
    --logging_steps="5000" \
    --save_steps="5000" \
    --eval_steps="5000" \
    --push_to_hub