Dear clients,
Meta Platforms Inc has announced last Friday it is releasing a large language model, the core software of a new AI system, to researchers, fueling the AI race as big tech companies rush to integrate the technology into their products and impress investors.
The battle for AI dominance began late last year with the launch of Microsoft-backed OpenAI ChatGPT and prompted tech heavyweights from Alphabet Inc to China’s Baidu Inc to pitch in with their solutions.
LLaMA, short for Large Language Model Meta AI, will be made available under a non-commercial license to researchers and organizations associated with government, civil society and academia, according to a promo blog.
The model, which Meta says requires „much less“ processing power than other alternatives, is trained in 20 languages, with a focus on Latin and Cyrillic languages.
AI has reinvigorated investments into the tech industry, whose slowdown has led to mass layoffs and cuts to experimental bets.
Meta claims that LLaMA outperforms competitors who explore more parameters or variables taken into account by the algorithm. In particular, the LLaMA version with 13 billion parameters could outperform GPT-3, the recent predecessor to the model on which ChatGPT is built.
