LlaMa 2: Everything You Need to Know About Meta’s New ChatGPT Rival

Meta has released an open-source version of its AI Model LLaMa 2, making it free for public use to compete with OpenAI’s GPT-4, which powers tools like ChatGPT. 

The Large Language Model (LLM), announced at Microsoft’s Inspire event, will allow developers, startups and businesses to build generative AI-enabled tools including AI chatbots like ChatGPT or Google’s Bard, and tweak them to their own purpose. 

The move comes as part of Meta’s “growing partnership” with Microsoft, with the two tech titans stating they share a “commitment to democratising AI and its benefits” by taking an open approach to AI developments.

“We believe an open approach is the right one for the development of today’s AI models,” the Facebook owner said in a blog post announcing the move. 

Opening access to today’s AI models means a generation of developers and researchers can stress test them, identifying and solving problems fast, as a community. By seeing how these tools are used by others, our own teams can learn from them, improve these tools, and fix vulnerabilities.

LlaMa 2 was trained on 40 per cent more data than its predecessor LLaMa 1, with new information scraped from “publicly available online data sources,” according to Meta.

This data includes tuning data from as recently as July 2023, allowing the model to offer more up-to-date information than its OpenAI counterpart – which has been trained on data up to September 2021. 

LLaMa is being will be free to use through Microsoft’s Azure platform, as well as AWS, Hugging Face and other providers, and will be free to use until an organisation releases 700 million monthly active users. 

'Safer and better' 

By making LlaMa 2 open-source, Meta says it wants to improve safety and transparency in the AI space. The company said that the LLaMa 2 model was “red-teamed for safety by generating adversarial prompts to facilitate model fine-tuning.” 

Nick Clegg, Meta’s president of global affairs, told BBC Radio 4’s Today programme on Wednesday that making LLMs open-source would also make them “safer and better” by inviting outside scrutiny.

“With the … wisdom of crowds you actually make these systems safer and better and, crucially, you take them out of the … clammy hands of the big tech companies which currently are the only companies that have either the computing power or the vast reservoirs of data to build these models in the first place.”

There is a possibility that by giving developers the chance to launch a rival to ChatGPT, Bard or Microsoft’s Bing chatbot, Meta is potentially breaking the competitive edge of industry giants such as Google.

LLaMa outperforms other LLMs like Falcon and MPT when it comes to reasoning, coding, proficiency, and knowledge tests. It is not quite at the level of GPT-4 just yet though, by Meta’s own admission. 

Microsoft is a key financial backer of OpenAI and has so far invested $11 billion into the research firm, but is nevertheless supporting the launch of LlaMa 2.

In a blog post of its own, the tech giant said the open-source AI model was just the “latest addition to our growing Azure model catalogue” 

“Azure’s purpose-built AI supercomputing platform is uniquely designed from the facility, hardware and software to support the world’s leading AI organizations to build, train and deploy some of the most demanding AI workloads. 

The availability of the Llama 2 models with Azure AI enables developers to take advantage of Azure AI’s powerful tooling for model training, fine-tuning, inference, and particularly the capabilities that support AI safety.

WeAreDevelopers World Congress 2023 brings together software developers and tech decision-makers from all around the globe together to discuss and learn about the latest trends and innovations shaping our world.

Taking place on 27-28 July 2023 in Berlin, Germany, the two-day event offers industry-leading speakers and an amazing community to exchange ideas, share knowledge, drive insightful conversations, facilitate networking and most of all - have fun.

As the world's leading event for software developers and tech experts, the event will provide a platform for everyone who wants to become a better professional to gain new skills and stay ahead of the curve.

Don't miss out on this incredible opportunity. REGISTER NOW!