Remove the support of old tfhub model that has list as inputs/outputs to reduce code complexity.

The new tfhub with dict as inputs/outputs have been released for a while, and users are expected to use new tfhub models.

PiperOrigin-RevId: 347314892

official/nlp/tasks/question_answering_test.py

@@ -21,8 +21,6 @@ import os
 from absl.testing import parameterized
 import tensorflow as tf
 
-from official.nlp.bert import configs
-from official.nlp.bert import export_tfhub
 from official.nlp.configs import bert
 from official.nlp.configs import encoders
 from official.nlp.data import question_answering_dataloader
@@ -133,26 +131,15 @@ class QuestionAnsweringTaskTest(tf.test.TestCase, parameterized.TestCase):
     self._run_task(config)
 
   def _export_bert_tfhub(self):
-    bert_config = configs.BertConfig(
-        vocab_size=30522,
-        hidden_size=16,
-        intermediate_size=32,
-        max_position_embeddings=128,
-        num_attention_heads=2,
-        num_hidden_layers=1)
-    _, encoder = export_tfhub.create_bert_model(bert_config)
-    model_checkpoint_dir = os.path.join(self.get_temp_dir(), "checkpoint")
-    checkpoint = tf.train.Checkpoint(model=encoder)
-    checkpoint.save(os.path.join(model_checkpoint_dir, "test"))
-    model_checkpoint_path = tf.train.latest_checkpoint(model_checkpoint_dir)
-
-    vocab_file = os.path.join(self.get_temp_dir(), "uncased_vocab.txt")
-    with tf.io.gfile.GFile(vocab_file, "w") as f:
-      f.write("dummy content")
-
+    encoder = encoders.build_encoder(
+        encoders.EncoderConfig(
+            bert=encoders.BertEncoderConfig(vocab_size=30522, num_layers=1)))
+    encoder_inputs_dict = {x.name: x for x in encoder.inputs}
+    encoder_output_dict = encoder(encoder_inputs_dict)
+    core_model = tf.keras.Model(
+        inputs=encoder_inputs_dict, outputs=encoder_output_dict)
     hub_destination = os.path.join(self.get_temp_dir(), "hub")
-    export_tfhub.export_bert_tfhub(bert_config, model_checkpoint_path,
-                                   hub_destination, vocab_file)
+    core_model.save(hub_destination, include_optimizer=False, save_format="tf")
     return hub_destination
 
   def test_task_with_hub(self):

official/nlp/tasks/sentence_prediction_test.py

@@ -21,8 +21,6 @@ from absl.testing import parameterized
 import numpy as np
 import tensorflow as tf
 
-from official.nlp.bert import configs
-from official.nlp.bert import export_tfhub
 from official.nlp.configs import bert
 from official.nlp.configs import encoders
 from official.nlp.data import sentence_prediction_dataloader
@@ -214,26 +212,15 @@ class SentencePredictionTaskTest(tf.test.TestCase, parameterized.TestCase):
     self.assertEqual(outputs["sentence_prediction"].shape.as_list(), [8, 1])
 
   def _export_bert_tfhub(self):
-    bert_config = configs.BertConfig(
-        vocab_size=30522,
-        hidden_size=16,
-        intermediate_size=32,
-        max_position_embeddings=128,
-        num_attention_heads=2,
-        num_hidden_layers=1)
-    _, encoder = export_tfhub.create_bert_model(bert_config)
-    model_checkpoint_dir = os.path.join(self.get_temp_dir(), "checkpoint")
-    checkpoint = tf.train.Checkpoint(model=encoder)
-    checkpoint.save(os.path.join(model_checkpoint_dir, "test"))
-    model_checkpoint_path = tf.train.latest_checkpoint(model_checkpoint_dir)
-
-    vocab_file = os.path.join(self.get_temp_dir(), "uncased_vocab.txt")
-    with tf.io.gfile.GFile(vocab_file, "w") as f:
-      f.write("dummy content")
-
+    encoder = encoders.build_encoder(
+        encoders.EncoderConfig(
+            bert=encoders.BertEncoderConfig(vocab_size=30522, num_layers=1)))
+    encoder_inputs_dict = {x.name: x for x in encoder.inputs}
+    encoder_output_dict = encoder(encoder_inputs_dict)
+    core_model = tf.keras.Model(
+        inputs=encoder_inputs_dict, outputs=encoder_output_dict)
     hub_destination = os.path.join(self.get_temp_dir(), "hub")
-    export_tfhub.export_bert_tfhub(bert_config, model_checkpoint_path,
-                                   hub_destination, vocab_file)
+    core_model.save(hub_destination, include_optimizer=False, save_format="tf")
     return hub_destination
 
   def test_task_with_hub(self):

official/nlp/tasks/tagging_test.py

@@ -20,8 +20,6 @@ import os
 import numpy as np
 import tensorflow as tf
 
-from official.nlp.bert import configs
-from official.nlp.bert import export_tfhub
 from official.nlp.configs import encoders
 from official.nlp.data import tagging_dataloader
 from official.nlp.tasks import tagging
@@ -98,26 +96,15 @@ class TaggingTest(tf.test.TestCase):
     task.initialize(model)
 
   def _export_bert_tfhub(self):
-    bert_config = configs.BertConfig(
-        vocab_size=30522,
-        hidden_size=16,
-        intermediate_size=32,
-        max_position_embeddings=128,
-        num_attention_heads=2,
-        num_hidden_layers=1)
-    _, encoder = export_tfhub.create_bert_model(bert_config)
-    model_checkpoint_dir = os.path.join(self.get_temp_dir(), "checkpoint")
-    checkpoint = tf.train.Checkpoint(model=encoder)
-    checkpoint.save(os.path.join(model_checkpoint_dir, "test"))
-    model_checkpoint_path = tf.train.latest_checkpoint(model_checkpoint_dir)
-
-    vocab_file = os.path.join(self.get_temp_dir(), "uncased_vocab.txt")
-    with tf.io.gfile.GFile(vocab_file, "w") as f:
-      f.write("dummy content")
-
+    encoder = encoders.build_encoder(
+        encoders.EncoderConfig(
+            bert=encoders.BertEncoderConfig(vocab_size=30522, num_layers=1)))
+    encoder_inputs_dict = {x.name: x for x in encoder.inputs}
+    encoder_output_dict = encoder(encoder_inputs_dict)
+    core_model = tf.keras.Model(
+        inputs=encoder_inputs_dict, outputs=encoder_output_dict)
     hub_destination = os.path.join(self.get_temp_dir(), "hub")
-    export_tfhub.export_bert_tfhub(bert_config, model_checkpoint_path,
-                                   hub_destination, vocab_file)
+    core_model.save(hub_destination, include_optimizer=False, save_format="tf")
     return hub_destination
 
   def test_task_with_hub(self):

official/nlp/tasks/utils.py

@@ -16,7 +16,6 @@
 """Common utils for tasks."""
 from typing import Any, Callable
 
-from absl import logging
 import orbit
 import tensorflow as tf
 import tensorflow_hub as hub
@@ -43,23 +42,7 @@ def get_encoder_from_hub(hub_model_path: str) -> tf.keras.Model:
       input_word_ids=input_word_ids,
       input_mask=input_mask,
       input_type_ids=input_type_ids)
-
-  # The legacy hub model takes a list as input and returns a Tuple of
-  # `pooled_output` and `sequence_output`, while the new hub model takes dict
-  # as input and returns a dict.
-  # TODO(chendouble): Remove the support of legacy hub model when the new ones
-  # are released.
-  hub_output_signature = hub_layer.resolved_object.signatures[
-      'serving_default'].outputs
-  if len(hub_output_signature) == 2:
-    logging.info('Use the legacy hub module with list as input/output.')
-    pooled_output, sequence_output = hub_layer(
-        [input_word_ids, input_mask, input_type_ids])
-    output_dict['pooled_output'] = pooled_output
-    output_dict['sequence_output'] = sequence_output
-  else:
-    logging.info('Use the new hub module with dict as input/output.')
-    output_dict = hub_layer(dict_input)
+  output_dict = hub_layer(dict_input)
 
   return tf.keras.Model(inputs=dict_input, outputs=output_dict)