internal change

PiperOrigin-RevId: 338095907

Author: aichendouble

official/nlp/tasks/question_answering.py

@@ -22,7 +22,6 @@
 import dataclasses
 import orbit
 import tensorflow as tf
-import tensorflow_hub as hub
 
 from official.core import base_task
 from official.core import config_definitions as cfg
@@ -87,11 +86,8 @@ def build_model(self):
       raise ValueError('At most one of `hub_module_url` and '
                        '`init_checkpoint` can be specified.')
     if self.task_config.hub_module_url:
-      hub_module = hub.load(self.task_config.hub_module_url)
-    else:
-      hub_module = None
-    if hub_module:
-      encoder_network = utils.get_encoder_from_hub(hub_module)
+      encoder_network = utils.get_encoder_from_hub(
+          self.task_config.hub_module_url)
     else:
       encoder_network = encoders.build_encoder(self.task_config.model.encoder)
     encoder_cfg = self.task_config.model.encoder.get()

official/nlp/tasks/question_answering_test.py

@@ -104,6 +104,7 @@ def _run_task(self, config):
     logs = task.aggregate_logs(step_outputs=logs)
     metrics = task.reduce_aggregated_logs(logs)
     self.assertIn("final_f1", metrics)
+    model.save(os.path.join(self.get_temp_dir(), "saved_model"))
 
   @parameterized.parameters(
       itertools.product(

official/nlp/tasks/sentence_prediction.py

@@ -23,7 +23,6 @@
 from scipy import stats
 from sklearn import metrics as sklearn_metrics
 import tensorflow as tf
-import tensorflow_hub as hub
 
 from official.core import base_task
 from official.core import config_definitions as cfg
@@ -77,11 +76,8 @@ def build_model(self):
       raise ValueError('At most one of `hub_module_url` and '
                        '`init_checkpoint` can be specified.')
     if self.task_config.hub_module_url:
-      hub_module = hub.load(self.task_config.hub_module_url)
-    else:
-      hub_module = None
-    if hub_module:
-      encoder_network = utils.get_encoder_from_hub(hub_module)
+      encoder_network = utils.get_encoder_from_hub(
+          self.task_config.hub_module_url)
     else:
       encoder_network = encoders.build_encoder(self.task_config.model.encoder)
     encoder_cfg = self.task_config.model.encoder.get()

official/nlp/tasks/sentence_prediction_test.py

@@ -86,6 +86,7 @@ def _run_task(self, config):
     iterator = iter(dataset)
     optimizer = tf.keras.optimizers.SGD(lr=0.1)
     task.train_step(next(iterator), model, optimizer, metrics=metrics)
+    model.save(os.path.join(self.get_temp_dir(), "saved_model"))
     return task.validation_step(next(iterator), model, metrics=metrics)
 
   @parameterized.named_parameters(

official/nlp/tasks/tagging.py

@@ -22,7 +22,6 @@
 from seqeval import metrics as seqeval_metrics
 
 import tensorflow as tf
-import tensorflow_hub as hub
 
 from official.core import base_task
 from official.core import config_definitions as cfg
@@ -89,11 +88,8 @@ def build_model(self):
       raise ValueError('At most one of `hub_module_url` and '
                        '`init_checkpoint` can be specified.')
     if self.task_config.hub_module_url:
-      hub_module = hub.load(self.task_config.hub_module_url)
-    else:
-      hub_module = None
-    if hub_module:
-      encoder_network = utils.get_encoder_from_hub(hub_module)
+      encoder_network = utils.get_encoder_from_hub(
+          self.task_config.hub_module_url)
     else:
       encoder_network = encoders.build_encoder(self.task_config.model.encoder)
 

official/nlp/tasks/tagging_test.py

@@ -73,6 +73,7 @@ def _run_task(self, config):
     optimizer = tf.keras.optimizers.SGD(lr=0.1)
     task.train_step(next(iterator), model, optimizer, metrics=metrics)
     task.validation_step(next(iterator), model, metrics=metrics)
+    model.save(os.path.join(self.get_temp_dir(), "saved_model"))
 
   def test_task(self):
     # Saves a checkpoint.

official/nlp/tasks/utils.py

@@ -22,11 +22,11 @@
 import tensorflow_hub as hub
 
 
-def get_encoder_from_hub(hub_model) -> tf.keras.Model:
+def get_encoder_from_hub(hub_model_path: str) -> tf.keras.Model:
   """Gets an encoder from hub.
 
   Args:
-    hub_model: A tfhub model loaded by `hub.load(...)`.
+    hub_model_path: The path to the tfhub model.
 
   Returns:
     A tf.keras.Model.
@@ -37,7 +37,7 @@ def get_encoder_from_hub(hub_model) -> tf.keras.Model:
       shape=(None,), dtype=tf.int32, name='input_mask')
   input_type_ids = tf.keras.layers.Input(
       shape=(None,), dtype=tf.int32, name='input_type_ids')
-  hub_layer = hub.KerasLayer(hub_model, trainable=True)
+  hub_layer = hub.KerasLayer(hub_model_path, trainable=True)
   output_dict = {}
   dict_input = dict(
       input_word_ids=input_word_ids,
@@ -49,6 +49,7 @@ def get_encoder_from_hub(hub_model) -> tf.keras.Model:
   # as input and returns a dict.
   # TODO(chendouble): Remove the support of legacy hub model when the new ones
   # are released.
+  hub_model = hub.load(hub_model_path)
   hub_output_signature = hub_model.signatures['serving_default'].outputs
   if len(hub_output_signature) == 2:
     logging.info('Use the legacy hub module with list as input/output.')