Yandex Creates Big Language Model For AI, Claims To Be Bigger Than Meta Platform Inc's Own.
Yandex creates a big language for AI. (photo: twitter @yandex)

JAKARTA - Russian technology company Yandex said on Thursday, June 23 that it has made a large language model for artificial intelligence research open to the public. Yandex hopes to spawn faster and deeper development of certain AI technologies.

Big language models, which have become a major trend in AI, are powerful programs that can generate paragraphs of text and imitate human conversations.

Yandex, like many Russian companies, has been in turmoil for several months since Moscow sent tens of thousands of troops to Ukraine on February 24.

The company, whose lines of business range from online search to e-commerce and ride-hailing services, suffered heavy losses in the first quarter of this year. These losses have flagged the possible negative consequences of sanctions against Russia. Yandex has also lost its chief executive and deputy chief executive, who stepped down as a form of responsibility for the loss.

Meta Platforms Inc on Wednesday, June 22, also released what it said was the largest open source artificial intelligence model to date. There is even a 66 billion parameter language model, after making the 175 billion parameter model available to the AI research community last month.

Yandex says its public model is bigger than what Meta claims.

"YaLM 100B contains 100 billion parameters, more than a similar GPT model available in open source under a license that allows commercial use," Yandex said, as quoted by Reuters.

"By making the YaLM 100B publicly available, we hope to provide an impetus to further develop generative neural networks," said Petr Popov, CEO of Yandex Technologies.

AI technology, is now a key area of research and development for several major online platforms. AI can perpetuate human social bias around issues such as race and gender, and some researchers have concerns about the dangers that large language models can spread.


The English, Chinese, Japanese, Arabic, and French versions are automatically generated by the AI. So there may still be inaccuracies in translating, please always see Indonesian as our main language. (system supported by DigitalSiber.id)