Farhan Naqvi

Write

Building a Distributed Storage Metadata Service in Go

Jun 13, 20266 min read

How are tokens processed by ML models?

Apr 28, 20242 min read

Building a Visual Workflow Designer for LLM Multi-Agent Systems

This project is a prototype that brings together visual workflow composition, agent orchestration, retrieval-augmented generation (RAG), and runtime inspection in a single experience.

Jun 15, 20264 min read

What are special tokens in Tokenization?

Apr 17, 20243 min read

What is Tokenization?

Apr 15, 20244 min read

How are context windows even relevant for LLMS?

In the context of LLMs, the concept of a "context window" refers to the span of tokens or words that the model considers when predicting the next word in a sequence of text. Here's how it works: Tokenization: Text input is tokenized into a sequence ...

Apr 8, 20242 min read

Attention Mechanisms

My previous article talks about the usage of attention mechanisms in order to mitigate token limits issue in RAG. It is a topic that many may not be aware of, so thought of covering it in one of my blogs. Attention mechanisms are a crucial component ...

Apr 5, 20244 min read

Strategies for Mitigating Token Limits Issue in RAG

Strategies for mitigating token limits issue involve techniques to enable large language models (LLMs) to focus on relevant parts of text when processing lengthy documents. Hierarchical Attention Mechanisms This technique allows the LLM to focus atte...

Apr 4, 20242 min read