Added some alternative code. About to do dataset

official/resnet/keras/keras_common.py

@@ -105,7 +105,8 @@ def get_optimizer():
     learning_rate = BASE_LEARNING_RATE * FLAGS.batch_size / 256
     optimizer = tf.train.MomentumOptimizer(learning_rate=learning_rate, momentum=0.9)
   else:
-    optimizer = tf.keras.optimizers.SGD(learning_rate=0.1, momentum=0.9)
+    # optimizer = tf.keras.optimizers.SGD(learning_rate=0.1, momentum=0.9)
+    optimizer = gradient_descent_v2.SGD(learning_rate=0.1, momentum=0.9)
 
   return optimizer
 

official/resnet/keras/keras_imagenet_main.py

@@ -23,6 +23,7 @@ from absl import flags
 import tensorflow as tf  # pylint: disable=g-bad-import-order
 
 from official.resnet import imagenet_main
+from official.resnet import imagenet_preprocessing
 from official.resnet import resnet_run_loop
 from official.resnet.keras import keras_common
 from official.resnet.keras import resnet50
@@ -30,6 +31,8 @@ from official.utils.flags import core as flags_core
 from official.utils.logs import logger
 from official.utils.misc import distribution_utils
 
+# import os
+# os.environ['TF2_BEHAVIOR'] = 'enabled'
 
 LR_SCHEDULE = [    # (multiplier, epoch to start) tuples
     (1.0, 5), (0.1, 30), (0.01, 60), (0.001, 80)
@@ -69,11 +72,25 @@ def learning_rate_schedule(current_epoch, current_batch, batches_per_epoch, batc
 
 def parse_record_keras(raw_record, is_training, dtype):
   """Adjust the shape of label."""
+  image_buffer, label, bbox = imagenet_main._parse_example_proto(raw_record)
+
+  image = imagenet_preprocessing.preprocess_image(
+      image_buffer=image_buffer,
+      bbox=bbox,
+      output_height=imagenet_main.DEFAULT_IMAGE_SIZE,
+      output_width=imagenet_main.DEFAULT_IMAGE_SIZE,
+      num_channels=imagenet_main.NUM_CHANNELS,
+      is_training=is_training)
+  image = tf.cast(image, dtype)
+  label = tf.sparse_to_dense(label, (imagenet_main.NUM_CLASSES,), 1)
+
+  """
   image, label = imagenet_main.parse_record(raw_record, is_training, dtype)
   # Subtract one so that labels are in [0, 1000), and cast to float32 for
   # Keras model.
   label = tf.cast(tf.cast(tf.reshape(label, shape=[1]), dtype=tf.int32) - 1,
       dtype=tf.float32)
+  """
   return image, label