small change

examples/ImageNetModels/vgg16.py

@@ -16,7 +16,7 @@ from imagenet_utils import (
     ImageNetModel, get_imagenet_dataflow, fbresnet_augmentor)
 
 
-def GroupNorm(x, group):
+def GroupNorm(x, group, gamma_initializer=tf.constant_initializer(1.)):
     """
     https://arxiv.org/abs/1803.08494
     """
@@ -39,7 +39,7 @@ def GroupNorm(x, group):
     beta = tf.get_variable('beta', [chan], initializer=tf.constant_initializer())
     beta = tf.reshape(beta, new_shape)
 
-    gamma = tf.get_variable('gamma', [chan], initializer=tf.constant_initializer(1.0))
+    gamma = tf.get_variable('gamma', [chan], initializer=gamma_initializer)
     gamma = tf.reshape(gamma, new_shape)
 
     out = tf.nn.batch_normalization(x, mean, var, beta, gamma, 1e-5, name='output')

examples/ResNet/imagenet-resnet.py

@@ -64,12 +64,11 @@ def get_config(model, fake=False):
 
     logger.info("Running on {} towers. Batch size per tower: {}".format(nr_tower, batch))
     if fake:
-        dataset_train = FakeData(
-            [[batch, 224, 224, 3], [batch]], 1000, random=False, dtype='uint8')
+        data = QueueInput(FakeData(
+            [[batch, 224, 224, 3], [batch]], 1000, random=False, dtype='uint8'))
         callbacks = []
     else:
-        dataset_train = get_data('train', batch)
-        dataset_val = get_data('val', batch)
+        data = QueueInput(get_data('train', batch))
 
         START_LR = 0.1
         BASE_LR = START_LR * (args.batch / 256.0)
@@ -87,6 +86,7 @@ def get_config(model, fake=False):
 
         infs = [ClassificationError('wrong-top1', 'val-error-top1'),
                 ClassificationError('wrong-top5', 'val-error-top5')]
+        dataset_val = get_data('val', batch)
         if nr_tower == 1:
             # single-GPU inference with queue prefetch
             callbacks.append(InferenceRunner(QueueInput(dataset_val), infs))
@@ -97,7 +97,7 @@ def get_config(model, fake=False):
 
     return TrainConfig(
         model=model,
-        dataflow=dataset_train,
+        data=data,
         callbacks=callbacks,
         steps_per_epoch=100 if args.fake else 1280000 // args.batch,
         max_epoch=105,

tensorpack/graph_builder/utils.py

@@ -330,6 +330,7 @@ class GradientPacker(object):
         if len(dtypes) != 1:
             logger.info("Skip GradientPacker due to inconsistent gradient types.")
             return False
+        self._grad_dtype = grads[0].dtype
 
         split_size = self._total_size // self._num_split
         split_size_last = self._total_size - split_size * (self._num_split - 1)
@@ -352,12 +353,14 @@ class GradientPacker(object):
 
         with cached_name_scope("GradientPacker", top_level=False):
             concat_grads = tf.concat([tf.reshape(g, [-1]) for g in grads], 0, name='concatenated_grads')
+            # concat_grads = tf.cast(concat_grads, tf.float16)
             grad_packs = tf.split(concat_grads, self._split_sizes)
             return grad_packs
 
     def unpack(self, grad_packs):
         with cached_name_scope("GradientPacker", top_level=False):
             concat_grads = tf.concat(grad_packs, 0, name='concatenated_packs')
+            # concat_grads = tf.cast(concat_grads, self._grad_dtype)
             flattened_grads = tf.split(concat_grads, self._sizes)
             grads = [tf.reshape(g, shape) for g, shape in zip(flattened_grads, self._shapes)]
             return grads