change to:

[Surya20794]

change to:
initial_state = dec_cell.zero_state(batch_size=batch_size,dtype=tf.float32).clone(cell_state=enc_state)

also

inference_logits, _ ,_ = tf.contrib.seq2seq.dynamic_decode(inference_decoder,
output_time_major=False,
impute_finished=True,
maximum_iterations=max_target_length)

and

training_logits, _ ,_ = tf.contrib.seq2seq.dynamic_decode(training_decoder,
output_time_major=False,
impute_finished=True,
maximum_iterations=max_target_length)

Originally posted by @Sofwath in #3 (comment)

[Surya20794]

training_logits= tf.contrib.seq2seq.dynamic_decode(training_decoder,

NameError: global name 'training_decoder' is not defined