Yandex Released A Bilingual Version Of Balaboba

Yandex has launched a bilingual version of the Balaboba text generator – now it supports both Russian and English. Balaboba demonstrates the capabilities of Yandex’s YaLM family of language models, which the company uses in more than 20 services, including Search and Alice.

The user just needs to write one or two words in Russian or English and choose one of the styles, and Balaboba will create a meaningful text on any topic, similar to the texts from the Internet, on which the model studied.

Balaboba can write a short story, and come up with a recipe, instructions or folk wisdom. And if you enter the name of the film, Balaboba will write a plot for it. Texts can be used as you wish. For example, describe products in an online store, find inspiration and ideas for advertising, or simply send interesting examples to friends on social networks.

Balaboba generates texts using the Yandex YaLM language model, which solves problems related to natural language processing. For example, YaLM models help Alice keep the conversation going, determine the topics of questions in Q’s, improve descriptions of orders on Services, and generate cards for quick answers in Search. Also, YaLM language models look for key points in the video, and generate advertisements and site descriptions (snippets).

To memorize the rules of the language and select the appropriate words for Balaboba, the parameters embedded inside the model allow, which change depending on whether the word is predicted correctly or incorrectly. The family of YaLM language models can have from 1 billion to 100 billion parameters.

Yandex recently released the largest bilingual model YaLM 100B with 100 billion parameters to the public. Balaboba uses its light version with 3 billion parameters. The model was trained on terabytes of texts of the English-language and Russian-language Internet in equal proportions.

You can try Balaboba here.

 

Dan Taylor
Dan Taylor is an experienced SEO consultant and has worked with brands and companies on optimizing for Russia (and Yandex) for a number of years. Winner of the inaugural 2018 TechSEO Boost competition, webmaster at HreflangChecker.com and Sloth.Cloud, and founder of RussianSearchNews.com.