Meta, the company formerly known as Facebook, has open-sourced LLAMA 2, its advanced AI model. Zuck might be onto something here. Let’s explore what LLAMA 2 brings to the table, its potential applications, and the implications of this open-source move for the AI community at large.
First and foremost, if you’re not a developer or an engineer, there’s nothing to see here. You really can’t use LLAMA 2 out of the box. It’s not an application like ChatGPT or Bard. LLAMA 2 is a large language model (like GPT-4, the underlying technology behind ChatGPT). It will be used to “power” applications.
That said, what makes the release of LLAMA 2 noteworthy is that it is open-source and available for both research and commercial use. Which means that developers and engineers can incorporate this very capable set of AI models into their own applications for free, which is remarkably different from the financial terms and conditions of companies such as OpenAI or Anthropic. Getting access to this kind of AI power for free is a big deal. There is one small catch: if you have “greater than 700 million monthly active users in the preceding calendar month,” you must request special permission from Meta to use LLAMA 2. Since most people in the business can name every company with more than 700 million monthly users off the top of their heads, this is not going to be much of an issue.
Meta has made LLAMA 2 available in three sizes: 7B, 13B, and 70B parameters. These models are trained on a mix of publicly available online data, making them versatile and adaptable to a wide range of tasks such as text generation, text completion, and question answering.
LLAMA 2 can be accessed through various platforms, including Hugging Face, AWS, and Azure. These platforms, widely used by developers and researchers, will provide well-known tooling for easy integration and management of LLAMA 2.
On the commercial side, Meta and Microsoft have a long-standing AI partnership that began with a joint effort to integrate ONNX Runtime with PyTorch, which enhanced the developer experience for PyTorch on Azure. This collaboration also saw Meta selecting Azure as a strategic cloud provider. The announcement of LLAMA 2’s availability on Azure is a continuation of this partnership, and it highlights Microsoft’s plans to expand its open-model ecosystem.
In the words of Yann LeCun, chief AI scientist at Meta, “This is going to change the landscape of the LLM market.” Indeed, the release of LLAMA 2 is newsworthy, and it will be fascinating to see how this open-source model shapes the future of AI.
And, as always, if you’re interested in learning how you can increase productivity with LLMs and generative AI, sign up for our free online course Generative AI for Execs.
Author’s note: This is not a sponsored post. I am the author of this article and it expresses my own opinions. I am not, nor is my company, receiving compensation for it.