Actually return the value from train_and_evaluate.
authorA. Unique TensorFlower <gardener@tensorflow.org>
Tue, 22 May 2018 18:44:52 +0000 (11:44 -0700)
committerTensorFlower Gardener <gardener@tensorflow.org>
Tue, 22 May 2018 18:47:35 +0000 (11:47 -0700)
PiperOrigin-RevId: 197597953

tensorflow/python/estimator/training.py

index dc5edc7..4f90bcf 100644 (file)
@@ -444,7 +444,7 @@ def train_and_evaluate(estimator, train_spec, eval_spec):
         'For distributed training, there can only be one `evaluator` task '
         '(with task id 0).  Given task id {}'.format(config.task_id))
 
-  executor.run()
+  return executor.run()
 
 
 class _StopAtSecsHook(session_run_hook.SessionRunHook):