Improving the speed of the response of the server


Presently I am using the docker image server ‘opennmt/opennmt:latest’ as ENG-GER
translation server. The response time is not good. Please find the below response time:

Server Running on Port 7784:
administrator:~/OpenNMT$ sudo docker run -v $PWD/data:/data -p 7784:7784 --entrypoint th opennmt/opennmt:latest tools/rest_translation_server.lua -host -model /data/data_en_de/model-ende_epoch7_5.55_release.t7 -bpe_model /data/data_en_de/train-ende.bpe32000 -replace_unk true -mode aggressive -segment_case -joiner_annotate
tput: No value for $TERM and no -T specified
**[09/06/19 09:45:27 INFO] Loading model **
**[09/06/19 09:45:27 INFO] Loading ‘/data/data_en_de/model-ende_epoch7_5.55_release.t7’… **
**[09/06/19 09:45:30 INFO] Model seq2seq trained on bitext **
**[09/06/19 09:45:30 INFO] Launch server **
Xavante started on port(s) 7784

*Client : *
administrator:~/OpenNMT$ python

Enter the sentence ==> hello world

Time taken : 1.1980292797088623

Do you want to continue: y or n ?y

Enter the sentence ==> I am trying to train the model using the following command:

Ich versuche, das Modell mit dem folgenden Befehl auszubilden:
Time taken : 5.142439603805542

Do you want to continue: y or n ?

How could we improve the response time from the server ?

Thank You,


You can try using a GPU. All information to run Docker with GPU support is here: