Decoding Llama3: An explainer for tinkerers
A not-so-quick 7-part guide to using the Llama3 open source AI model
The arrival of the Llama3 large language model in April 2024 was an exciting next step for those of us interested in exlopring the potential of AI and in experimenting in how we can apply it to a variety of applications.
In this 7-part guide for Llama3, I have tried to describe how the model can be utilised and the ways in which we can explore its new reasoning capabilites and improved performance compared to previous generation LLMs. The appeal for engineers of course, is the significant performance boost and its open source licensing policy.
To get started with reading this 7 part guide, use the link below
Contents: Part 1 to 7 of this Decoding Llama3 guide
Part 1 - Intro to Llama3 <START READING HERE
Part 2 - Understanding the configuration
Part 3 - Normalisation
Part 4 - Rotary Positional Embeddings
Part 5 - Grouped Query Attention
Part 6 - Feed Forward Network
Part 7 - TransformerBlock
Hosted by
NLP specialist with 14 years of experience building products. Exploring productivity hacks.
Supported by
