Fix use crossentropy_onehot_grad instead of logsoftmax

theano/tensor/nnet/nnet.py

@@ -678,7 +678,15 @@ def local_logsoftmax_grad(node):
             len(node.inputs) == 2 and
             isinstance(node.inputs[0].owner.op, tensor.Elemwise) and
             node.inputs[0].owner.inputs[1].owner.op == Softmax() and
-                node.inputs[1] == node.inputs[0].owner.inputs[1]):
+            node.inputs[1] == node.inputs[0].owner.inputs[1]) and not(
+                # skip if it will be optimized by
+                # local_advanced_indexing_crossentropy_onehot_grad
+                node.inputs[1].owner and node.inputs[1].owner.op in
+                (softmax_op, softmax_with_bias) and
+                node.inputs[0].owner.op == tensor.true_div and
+                node.inputs[0].owner.inputs[1] == node.inputs[1] and
+                isinstance(node.inputs[0].owner.inputs[0].owner.op,
+                           subtensor.AdvancedIncSubtensor)):
             # get parameters from unoptimized op
             sm = node.inputs[0].owner.inputs[1]
             # sm_input = node.inputs[1].owner.inputs[0]
@@ -1568,7 +1576,7 @@ def local_advanced_indexing_crossentropy_onehot(node):
             sm = log.owner.inputs[0]
 
     # Second case: log(softmax(x)[rows, labels])
-    if node.op == tensor.log:
+    elif node.op == tensor.log:
         pre_log = node.inputs[0].owner
         if pre_log and isinstance(pre_log.op, subtensor.AdvancedSubtensor):
             try: