Meta recently released two large-language models, Llama and Llama 2.
Some of the scientists and engineers who worked on Llama have left Meta and joined other companies or AI startups.
Meta CEO Mark Zuckerberg launched its own generative AI bot, called LLaMA or the Large Language Model Meta AI, earlier this year. Following the release of the first-generation LLaMA, Meta, the parent company of Facebook, collaborated with Microsoft to launch the advanced Llama 2.
However, after introducing two AI models in less than a year, the social media giant is reportedly facing a major problem. According to a report from The Information, some of the scientists and engineers who worked on Llama have left Meta and joined other companies or AI startups.
Meta recently released two large-language models, Llama and Llama 2, which received praise for being free and open-source alternatives to models from OpenAI and Anthropic. However, some of the scientists and engineers who contributed to Llama felt that the recognition came too late, the report said.
The report suggests that a dispute over access to computing resources within the company led to this situation. This “previously unreported” competition forced the “embittered” employees to quit.
“Many have quit, embittered by a previously unreported internal battle over computing resources with another Meta research team working on a rival model that the company ultimately abandoned,” the report said.
Released In July, Llama 2 is a successor of Llama, which was capable of generating text and code like other chatbot systems. It is available for free for both research and commercial purposes.
Llama 2 is a collection of models that could generate text and code in response to prompts, comparable to other chatbot-like systems. It is a large language model (LLM) that is more powerful and efficient than previous models. It is trained on a massive dataset of text and code and can generate more creative and informative text than Google’sBard or ChatGPT.
Meta has trained Llama 2 using a combination of publicly accessible data. According to Meta, it shows significant improvements in performance compared to the previous generation of Llama models.
