run_train.sh

python run_mlm_flax.py \
    --output_dir="./" \
    --model_type="bert" \
    --hub_model_id="norwegian-bert-large-newcode" \
    --config_name="./" \
    --tokenizer_name="./" \
    --dataset_name="oscar" \
    --dataset_config_name="unshuffled_deduplicated_no" \
    --max_seq_length="128" \
    --weight_decay="0.01" \
    --per_device_train_batch_size="8" \
    --per_device_eval_batch_size="2" \
    --learning_rate="3e-4" \
    --warmup_steps="1000" \
    --overwrite_output_dir \
    --num_train_epochs="20" \
    --adam_beta1="0.9" \
    --adam_beta2="0.98" \
    --logging_steps="500" \
    --save_steps="2500" \
    --eval_steps="100000000" \
    --dtype="bfloat16" \
    --trust_remote_code \
    --push_to_hub