- Meta has unveiled Llama 3.1 405B, its largest open source AI model to date, boasting 405 billion parameters, making it competitive with leading proprietary models like OpenAI’s GPT-4o and Anthropic’s Claude 3.5 Sonnet.
- Trained using 16,000 Nvidia H100 GPUs and a dataset of 15 trillion tokens, Llama 3.1 405B is designed to handle a variety of text-based tasks, from coding to summarizing documents in multiple languages.
- Meta is also exploring multimodality, developing Llama models that can recognize images and videos, and understand speech, although these are not yet ready for public release.
- The company has updated Llama’s license to allow developers to use outputs from the Llama 3.1 model family to develop third-party AI generative models, aiming to foster an ecosystem around its AI tools.
- Despite its advancements, Llama 3.1 405B requires significant hardware to run and Meta continues to face scrutiny over its use of training data, including copyrighted materials.
Source: Techcrunch
