fix

formath · formath · commit 860ef4ed1f12 · 2017-10-23T15:22:33.000+08:00
diff --git a/python/train.py b/python/train.py
@@ -91,7 +91,6 @@
 train_op = optimizer.minimize(cost, global_step=global_step)
 
 # eval
-label_num = 2
 tf.get_variable_scope().reuse_variables()
 
 # train cross entropy loss
