Merge pull request #3 from tensorflow/master

Updated

official/vision/image_classification/resnet_runnable.py

@@ -20,6 +20,7 @@ from __future__ import print_function
 
 import tensorflow.compat.v2 as tf
 
+from official.modeling import performance
 from official.staging.training import standard_runnable
 from official.staging.training import utils
 from official.utils.flags import core as flags_core
@@ -85,21 +86,15 @@ class ResnetRunnable(standard_runnable.StandardTrainable,
     # Make sure iterations variable is created inside scope.
     self.global_step = self.optimizer.iterations
 
-    if self.dtype == tf.float16:
-      loss_scale = flags_core.get_loss_scale(flags_obj, default_for_fp16=128)
-      self.optimizer = (
-          tf.keras.mixed_precision.experimental.LossScaleOptimizer(
-              self.optimizer, loss_scale))
-    elif flags_obj.fp16_implementation == 'graph_rewrite':
-      # `dtype` is still float32 in this case. We built the graph in float32
-      # and let the graph rewrite change parts of it float16.
-      if not flags_obj.use_tf_function:
+    use_graph_rewrite = flags_obj.fp16_implementation == 'graph_rewrite'
+    if use_graph_rewrite and not flags_obj.use_tf_function:
       raise ValueError('--fp16_implementation=graph_rewrite requires '
                        '--use_tf_function to be true')
-      loss_scale = flags_core.get_loss_scale(flags_obj, default_for_fp16=128)
-      self.optimizer = (
-          tf.train.experimental.enable_mixed_precision_graph_rewrite(
-              self.optimizer, loss_scale))
+    self.optimizer = performance.configure_optimizer(
+        self.optimizer,
+        use_float16=self.dtype == tf.float16,
+        use_graph_rewrite=use_graph_rewrite,
+        loss_scale=flags_core.get_loss_scale(flags_obj, default_for_fp16=128))
 
     self.train_loss = tf.keras.metrics.Mean('train_loss', dtype=tf.float32)
     self.train_accuracy = tf.keras.metrics.SparseCategoricalAccuracy(

research/README.md

@@ -5,6 +5,10 @@ This folder contains machine learning models implemented by researchers in
 respective authors. To propose a model for inclusion, please submit a pull
 request.
 
+**Note: some research models are stale and have not updated to the latest
+TensorFlow yet. If users have trouble with TF 2.x for research models,
+please consider TF 1.15.**
+
 ## Models
 
 -   [adversarial_crypto](adversarial_crypto): protecting communications with