We pride ourselves in delivering consistent quality translations. This is not just what we say, it is what we do. Pangeanic recruits the best...
It has been a successful day at PangeaMT today and we completed the successful re-training of a large software-related Brazilian Portuguese engine. This is one of our largest engines with around 18 million unique segments (sentences) for training. It is so large that it even can translate into other fields not related to software when we upload short glossaries. The exact number of words is 229 million English words and 249 in Brazilian Portuguese This is a monster engine trained bidirectionally, that is into and out of English into and out of Brazilian Portuguese. With a BLEU over 65%, this engine covers practically everything you need or want to say in Brazilian Portuguese. That is some engine saving hundreds of hours and thousands of dollars and Euros in localization work, if you ask me...
Related Posts
Artificial intelligence model training is a process in which an algorithm is taught to correctly interpret data and make accurate decisions based on...
Pangeanic is currently developing an R+D project called “COR: FAST AND EFFICIENT TECHNICAL TRANSLATION MANAGEMENT”, having successfully overcome...