Inference with NMT

Understand how to implement inference in neural machine translation models by defining encoder-decoder setups for recursive prediction. Learn to trigger the decoder, feed predictions step-by-step, and generate translated text without target sequences during inference.