Instructions on how to integrate the trained model with your application are available on the model page, after the training is complete.
For a model to properly pick up syntax relationships, multiple phrases need to be provided.
Try gathering and providing more source text. The richer this is, the better your predictions will be.
Model training is a computational intensive task, we are allocating resources for it on the fly. A typical model should start training in a couple of minutes.
Try to make sure that you are providing only essential text data as your training source.
If your needs exceed the free plan, please reach out to us via the contact form.
There is no affinity towards a specific language. You should get successful results with most european based alphabets.
If you have found any issues with training on source text in your native language, please reach out to us via the contact form. Do not forget to provide a link to the model.