Meta has officially released Llama 3, the latest iteration of its open-source large language model. The new model promises significant improvements in performance, reasoning, and efficiency over its predecessor, Llama 2. Available in multiple sizes, Llama 3 is designed to empower developers and researchers with state-of-the-art AI capabilities while maintaining an open license.
According to Meta, Llama 3 has been trained on a diverse dataset, leading to better understanding and generation of natural language. Early benchmarks show it outperforms many comparable models in tasks like reading comprehension, mathematics, and code generation. The company emphasizes that the model is built with safety and responsibility in mind, incorporating advanced filtering and fine-tuning techniques.
The release includes both base and instruction-tuned versions, with sizes ranging from 8 billion to 70 billion parameters. Meta encourages the community to build upon Llama 3, fostering innovation in AI applications. This launch reinforces Meta's commitment to open science and democratizing access to powerful AI tools.