fixed the tf issue #74564

I looked into the issue and found out that after the evaluation we are getting a list of size 4 where only the last one is the accuracy of the export_model. So ignored all the other output to only get the required loss and accuracy

site/en/tutorials/keras/text_classification.ipynb

@@ -861,7 +861,7 @@
         ")\n",
         "\n",
         "# Test it with `raw_test_ds`, which yields raw strings\n",
-        "loss, accuracy = export_model.evaluate(raw_test_ds)\n",
+        "_, _, loss, accuracy = export_model.evaluate(raw_test_ds)\n",
         "print(accuracy)"
       ]
     },