Meta and Microsoft Introduce the Next Generation of Llama

Meta and Microsoft have jointly announced the expansion of their artificial intelligence partnership by introducing their new large language model (LLM) called “Llama 2“. This cutting-edge language model is now available for both research and commercial use, making it a potential competitor to ChatGPT (OpenAI) and Bard (Google). The offering of Llama 2 for free enhances its accessibility and usability in various domains.

Llama 2: Features

Llama 2 comes in three different sizes: 7B, 13B, and 70B.
This new large language model (LLM) demonstrates superior performance compared to other open-source LLMs across various external benchmarks, excelling in reasoning, coding proficiency, and knowledge tests.
Pre-trained on publicly available online data sources, the Llama-2-chat, which is a fine-tuned model, makes use of publicly available instruction datasets and benefits from over 1 million human annotations.

Llama 2: Multi-platform Accessibility and Seamless Integration

Llama 2 can be found in the Azure AI model catalog, empowering developers using Microsoft Azure to incorporate it into their projects and take advantage of cloud-native tools for content filtering and safety features.
Additionally, Llama 2 is optimized for local execution on Windows, offering developers a smooth workflow while delivering generative AI experiences to users across various platforms. Moreover, Llama 2 is accessible through Amazon Web Services (AWS), Hugging Face, and other providers, widening its availability to a broader audience.