D3.2:

This deliverable provides details of Task 3.2, whose main objective is to train Neural Machine Translation (NMT) systems for the languages supported in the RAYUELA project: English, Spanish, Dutch, French, Greek, Estonian, and Slovak. 

In general lines, this deliverable describes how the NMT systems were developed, including data used for the specific translation systems and data processing steps for their development. 

Section 1 gives an overview of the development of the NMT systems and the workflows applied to process language data for NMT system development. Section 2 provides an overview of the NMT systems developed for the project purpose together with information on the size of the data used. Section 3 describes basic parameters for the API and the API key generated for the use in the project. Section 4 lists two examples: how partners can access NMT systems and where NMT systems have been integrated. Finally, section 5 highlights the main conclusions of the deliverable. 

D3.3: New features:

This deliverable provides details of Task 3.2, whose main objective is to train Neural Machine Translation (NMT) systems for the languages supported in the RAYUELA project: English, Spanish, Dutch, French, Greek, Estonian, and Slovak. 

This deliverable describes the progress made since delivering of the D3.2 RAYUELA machine translation module [1] which described how the NMT systems were developed, including data used for the specific translation systems and data processing steps for their development. This deliverable describes language coverage of the NMT engines, and how NMT system was used to prepare translations for use within the Serious Game developed and deployed by WP3. 

Section 1 gives an overview of the NMT systems available for the RAYUELA project partners translation tasks. Section 2 describes the basic parameters for the API and the API key generated for use in the project. Section 3 lists two examples: how partners can access NMT systems and where NMT systems have been integrated. Finally, section 4 highlights the main conclusions of this deliverable.