Update losses implementation

tensorflow/contrib/learn/python/learn/estimators/logistic_regressor_test.py

@@ -26,12 +26,12 @@
 from tensorflow.contrib.learn.python.learn.datasets import base
 from tensorflow.contrib.learn.python.learn.estimators import logistic_regressor
 from tensorflow.contrib.learn.python.learn.estimators import metric_key
-from tensorflow.contrib.losses.python.losses import loss_ops
 from tensorflow.python.framework import constant_op
 from tensorflow.python.framework import dtypes
 from tensorflow.python.ops import array_ops
 from tensorflow.python.ops import init_ops
 from tensorflow.python.ops import math_ops
+from tensorflow.python.ops.losses import losses
 from tensorflow.python.platform import test
 
 
@@ -55,7 +55,7 @@ def _logistic_regression_model_fn(features, labels, mode):
       # AUC/precision/recall/etc will change meaningfully even on a toy dataset.
       biases_initializer=init_ops.constant_initializer(-10.0))
   predictions = math_ops.sigmoid(logits)
-  loss = loss_ops.sigmoid_cross_entropy(logits, labels)
+  loss = losses.sigmoid_cross_entropy(labels, logits)
   train_op = optimizers.optimize_loss(
       loss, variables.get_global_step(), optimizer='Adagrad', learning_rate=0.1)
   return predictions, loss, train_op

tensorflow/contrib/learn/python/learn/ops/losses_ops.py

@@ -20,14 +20,14 @@
 from __future__ import print_function
 
 from tensorflow.contrib.framework import deprecated
-from tensorflow.contrib.losses.python.losses import loss_ops
 from tensorflow.python.framework import ops
 from tensorflow.python.ops import array_ops as array_ops_
 from tensorflow.python.ops import math_ops
 from tensorflow.python.ops import nn
+from tensorflow.python.ops.losses import losses
 
 
-@deprecated('2016-12-01', 'Use `tf.contrib.losses.mean_squared_error` '
+@deprecated('2016-12-01', 'Use `tf.losses.mean_squared_error` '
             'and explicit logits computation.')
 def mean_squared_error_regressor(tensor_in, labels, weights, biases, name=None):
   """Returns prediction and loss for mean squared error regression."""
@@ -36,10 +36,10 @@ def mean_squared_error_regressor(tensor_in, labels, weights, biases, name=None):
     predictions = nn.xw_plus_b(tensor_in, weights, biases)
     if len(labels.get_shape()) == 1 and len(predictions.get_shape()) == 2:
       predictions = array_ops_.squeeze(predictions, squeeze_dims=[1])
-    return predictions, loss_ops.mean_squared_error(predictions, labels)
+    return predictions, losses.mean_squared_error(labels, predictions)
 
 
-@deprecated('2016-12-01', 'Use `tf.contrib.losses.softmax_cross_entropy` '
+@deprecated('2016-12-01', 'Use `tf.losses.softmax_cross_entropy` '
             'and explicit logits computation.')
 def softmax_classifier(tensor_in,
                        labels,
@@ -72,4 +72,4 @@ def softmax_classifier(tensor_in,
     logits = nn.xw_plus_b(tensor_in, weights, biases)
     if class_weight is not None:
       logits = math_ops.multiply(logits, class_weight)
-    return nn.softmax(logits), loss_ops.softmax_cross_entropy(logits, labels)
+    return nn.softmax(logits), losses.softmax_cross_entropy(labels, logits)