While half of the IT people were excited about the Llama 3.1 release, you had to read the entire press release. As example, Meta’s Agentic System focuses on creating autonomous agents that can perform tasks, make decisions, and interact with their environments independently, emphasizing goal-oriented behavior and adaptive responses. 

Llama Stack

The question is what is this all?

“Llama Stack is a set of standardized and opinionated interfaces for how to build canonical toolchain components (fine-tuning, synthetic data generation) and agentic applications. Our hope is for these to become adopted across the ecosystem, which should help with easier interoperability.”

So it is a hole eco system :

• Real-time and batch inference
• Supervised fine-tuning
• Evaluation of your model for your specific application
• Continual pre-training
• Retrieval-Augmented Generation (RAG)
• Function calling
• Synthetic data generation

I’ve started building around this new ecosystem. It will be a complete open-source copilot, similar to the Gnoppix copilot, which will eventually be replaced by this new standard.

Visit Meta’s Github:

LLama Agentic System Installation and Setup Guide

Scripts for fine-tuning Meta Llama3 with composable FSDP & PEFT methods to cover single/multi-node GPUs