scenic/projects/baselines/bert/main.py

# Copyright 2025 The Scenic Authors.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

"""Main script for BERT."""
import importlib
from typing import Any

from absl import flags
from clu import metric_writers
import jax
import jax.numpy as jnp
import ml_collections
from scenic import app
from scenic.projects.baselines.bert import model as bert_model
from scenic.projects.baselines.bert import trainer as bert_trainer
from scenic.train_lib import train_utils


FLAGS = flags.FLAGS


def get_model_cls(model_name: str) -> Any:
  """Returns model class given its name."""
  if model_name == 'bert':
    return bert_model.BERTModel
  if model_name == 'bert_classification':
    return bert_model.BERTClassificationModel
  if model_name == 'bert_regression':
    return bert_model.BERTRegressionModel
  else:
    raise ValueError(f'Unrecognized model: {model_name}.')


def main(rng: jnp.ndarray, config: ml_collections.ConfigDict, workdir: str,
         writer: metric_writers.MetricWriter):
  """Main function for the BERT project."""
  if 'glue_task' in config:
    task_config_file = importlib.import_module(
        f'scenic.projects.baselines.bert.configs.glue.tasks.bert_{config.glue_task}_config'
    )
    config = task_config_file.get_config(
        variant=config.variant, init_from=config.init_from)
  # Build the loss_fn, metrics, and flax_model.
  model_cls = get_model_cls(config.model_name)
  data_rng, rng = jax.random.split(rng)
  dataset = train_utils.get_dataset(
      config, data_rng, dataset_service_address=FLAGS.dataset_service_address)
  bert_trainer.train(
      rng=rng,
      config=config,
      model_cls=model_cls,
      dataset=dataset,
      workdir=workdir,
      writer=writer)


if __name__ == '__main__':
  app.run(main=main)