./run_mlm_flax.py \
    --output_dir="./" \
    --model_type="roberta" \
    --model_name_or_path="./" \
    --config_name="./" \
    --tokenizer_name="./" \
    --train_file /mnt/disks/flaxdisk/corpus/train_4_4.json \
    --validation_file /mnt/disks/flaxdisk/corpus/validation.json \
    --cache_dir="/mnt/disks/flaxdisk/cache/" \
    --max_seq_length="512" \
    --weight_decay="0.01" \
    --per_device_train_batch_size="40" \
    --per_device_eval_batch_size="40" \
    --learning_rate="5e-5" \
    --end_learning_rate="0" \
    --warmup_steps="0" \
    --overwrite_output_dir \
    --num_train_epochs="2" \
    --adam_beta1="0.9" \
    --adam_beta2="0.98" \
    --adam_epsilon="1e-6" \
    --logging_steps="10000" \
    --save_steps="10000" \
    --eval_steps="10000" \
    --preprocessing_num_workers="64" \
    --auth_token="True" \
    --dtype="bfloat16" \
    --push_to_hub