training.py _slice_arrays() fix crash when arrays are None (keras-tea…

…m#7069)

* training.py _slice_arrays() fix crash when arrays are None

* training.py test _slice_arrays()

keras/engine/training.py

@@ -387,21 +387,25 @@ def _slice_arrays(arrays, start=None, stop=None):
     # Returns
         A slice of the array(s).
     """
-    if isinstance(arrays, list):
+    if arrays is None:
+        return [None]
+    elif isinstance(arrays, list):
         if hasattr(start, '__len__'):
             # hdf5 datasets only support list objects as indices
             if hasattr(start, 'shape'):
                 start = start.tolist()
-            return [x[start] for x in arrays]
+            return [None if x is None else x[start] for x in arrays]
         else:
-            return [x[start:stop] for x in arrays]
+            return [None if x is None else x[start:stop] for x in arrays]
     else:
         if hasattr(start, '__len__'):
             if hasattr(start, 'shape'):
                 start = start.tolist()
             return arrays[start]
-        else:
+        elif hasattr(start, '__getitem__'):
             return arrays[start:stop]
+        else:
+            return [None]
 
 
 def _weighted_masked_objective(fn):

tests/keras/engine/test_training.py

@@ -9,6 +9,7 @@
 from keras.engine.training import Model
 from keras.engine.training import _check_loss_and_target_compatibility
 from keras.engine.training import _weighted_masked_objective
+from keras.engine.training import _slice_arrays
 from keras.models import Sequential
 from keras import backend as K
 from keras.utils import Sequence
@@ -29,6 +30,27 @@ def __getitem__(self, idx):
             np.random.random((self.batch_size, 3))]
 
 
+@keras_test
+def test_slice_arrays():
+    input_a = np.random.random((10, 3))
+    _slice_arrays(None)
+    _slice_arrays(input_a, 0)
+    _slice_arrays(input_a, 0, 1)
+    _slice_arrays(input_a, stop=2)
+    input_a = [None, [1, 1], None, [1, 1]]
+    _slice_arrays(input_a, 0)
+    _slice_arrays(input_a, 0, 1)
+    _slice_arrays(input_a, stop=2)
+    input_a = [None]
+    _slice_arrays(input_a, 0)
+    _slice_arrays(input_a, 0, 1)
+    _slice_arrays(input_a, stop=2)
+    input_a = None
+    _slice_arrays(input_a, 0)
+    _slice_arrays(input_a, 0, 1)
+    _slice_arrays(input_a, stop=2)
+
+
 @keras_test
 def test_weighted_masked_objective():
     a = Input(shape=(3,), name='input_a')