LLama 3.2 Released

Meta just dropped Llama 3.2: AI at Meta Blog Post.

Here's some highlights from their blog post:

Vision and Lightweight Models: Llama 3.2 includes vision-enabled models (11B and 90B) designed for tasks like document understanding and image reasoning, along with lightweight text-only models (1B and 3B) optimized for mobile and edge devices. These models are available in pre-trained and instruction-tuned versions for easier customization.

Edge AI Capabilities: The lightweight models support a 128K token context length and are optimized for on-device processing, enabling real-time tasks like summarization and instruction following on hardware like Qualcomm, MediaTek, and Arm. This allows for faster responses and enhanced privacy by keeping data local.

Llama 3.2 models are available for download and immediate development.

My takeaways:

• Llama still has Meta's restrictive licensing. Don't use it blindly for your production workloads without consulting an expert on their license and your business.
• Llama 3.2's high token context length for the 1B and 3B models is impressive. This will enable some really cool use cases. Especially running on mobile and edge devices.
• It's great to see a new vision model (in 11B and 90B sizes).