t5_udf = udf(t5_seq2seq_udf, returnType=StringType())

results_df = input_df.withColumn('output_column', t5_udf(input_df['sentence']))

results_df.show(truncate=False)