| ./run_mlm_flax.py \ | |
| --output_dir="./" \ | |
| --model_type="roberta" \ | |
| --config_name="./" \ | |
| --tokenizer_name="./" \ | |
| --dataset_name="large_spanish_corpus" \ | |
| --dataset_config_name="combined" \ | |
| --max_seq_length="128" \ | |
| --per_device_train_batch_size="4" \ | |
| --per_device_eval_batch_size="4" \ | |
| --learning_rate="3e-4" \ | |
| --warmup_steps="1000" \ | |
| --overwrite_output_dir \ | |
| --num_train_epochs="8" \ | |
| --push_to_hub | |