Transformer based neural machine translation for high-performance inference
Created by: xptree
Hi, I notice that you have released Transformer based NMT model at https://github.com/PaddlePaddle/models/tree/develop/PaddleNLP/neural_machine_translation/transformer. Excellent work!!!
I want to try the current model in production which has high QPS requirement. I wonder if PaddlePaddle has any plan to release a Transformer based NMT model for high performance inference, just like what you have done for PaddleDetection (https://github.com/PaddlePaddle/models/pull/3529).
Any help will be appreciated.