In the early days of high-performance computing, each of the major tech corporations made significant investments in establishing their own closed-source Unix versions. It was difficult to believe at the time that any other approach could produce such powerful software. However, open source Linux eventually gained popularity, first because it allowed developers to tweak its code as they saw fit and was less expensive, and then because it became more powerful, more secure, and had a bigger ecosystem that supported more features than any closed Unix.
Today, Linux serves as the industry standard foundation for both cloud computing and the operating systems that power the majority of mobile devices, and as a result, we all benefit from improved goods.
He stated, “I believe AI will evolve in a similar manner. Currently, several tech businesses are creating industry-leading closed models, but open source is rapidly closing the gap. Last year, Llama 2 was akin to an older generation of vehicles trailing behind the Frontier. This year, Llama 3 competes with the most advanced models and even excels in some areas. Starting next year, we expect future Llama models to be the most advanced in the industry. Even now, Llama is setting the standard for openness, modifiability, and cost efficiency.
Today, we are taking the next steps towards making open source AI the industry norm. We’re releasing Llama 3.1 405B, the first frontier-level open source AI model, along with updated Llama 3.1 70B and 8B models. The 405B model’s open design, combined with its superior cost/performance ratio compared to closed versions, makes it the best choice for fine-tuning and distilling smaller models.
Beyond providing these models, we’re collaborating with a variety of companies to expand the broader ecosystem. Amazon, Databricks, and NVIDIA are introducing comprehensive suites of services to help developers fine-tune and streamline their own models. Innovators such as Groq have developed low-latency, low-cost inference services for all of the new models. These models will be available across all major clouds, including AWS, Azure, Google, Oracle, and others. Companies like Scale.AI, Dell, and Deloitte are ready to assist enterprises with implementing Llama and training custom models using their own data. As the community grows and more organizations build new services, we can work together to make Llama the industry standard, bringing the benefits of AI to everyone.”
Meta is committed to open-source AI. I’ll explain why I feel open source is the greatest development stack for you, why open sourcing Llama is good for Meta, and why open source AI is good for the world, making it a platform that will be there for a long time.
