Update 2. 改造后的mnist_train.ipynb

Author: perhapszzy

Deep_Learning_with_TensorFlow/1.0.0/Chapter09/2. 改造后的mnist_train.ipynb

@@ -99,7 +99,7 @@
     "                _, loss_value, step = sess.run(\n",
     "                    [train_op, loss, global_step], feed_dict={x: xs, y_: ys},\n",
     "                    options=run_options, run_metadata=run_metadata)\n",
-    "                writer.add_run_metadata(run_metadata=run_metadata, tag=(\"tag%d\"%i), global_step=i),
+    "                writer.add_run_metadata(run_metadata=run_metadata, tag=(\"tag%d\"%i), global_step=i),\n",
     "                print(\"After %d training step(s), loss on training batch is %g.\" % (step, loss_value))\n",
     "            else:\n",
     "                _, loss_value, step = sess.run([train_op, loss, global_step], feed_dict={x: xs, y_: ys})\n",