LoRA (Low-Rank Adaptation) is a technique for efficiently fine-tuning LLMs by introducing low-rank trainable weight matrices into specific model layers.

large_Male-Author.jpeg

medium_Male-Author.jpeg

small_Male-Author.jpeg

thumbnail_Male-Author.jpeg

Male-Author.jpeg

Ruben Winastwan, Freelance Technical Writer

Ruben Winastwan

large_LoRA_ Low-Rank Adaptation of Large Language Models-20241113-102322.png

medium_LoRA_ Low-Rank Adaptation of Large Language Models-20241113-102322.png

small_LoRA_ Low-Rank Adaptation of Large Language Models-20241113-102322.png

thumbnail_LoRA_ Low-Rank Adaptation of Large Language Models-20241113-102322.png

LoRA_ Low-Rank Adaptation of Large Language Models-20241113-102322.png

Efficient LLM Fine-tuning with LoRA (Low-Rank Adaptation)

LoRA Explained: Low-Rank Adaptation for Fine-Tuning LLMs

Robert Tran, the Co-founder and CTO of Inkeep, shared how Inkeep and Zilliz built an AI-powered assistant for their documentation site. 

large_How Inkeep and Zilliz built an AI assistant-20241023-061518.png

medium_How Inkeep and Zilliz built an AI assistant-20241023-061518.png

small_How Inkeep and Zilliz built an AI assistant-20241023-061518.png

thumbnail_How Inkeep and Zilliz built an AI assistant-20241023-061518.png

How Inkeep and Zilliz built an AI assistant-20241023-061518.png

Inkeep Built an Efficient RAG-driven AI Assistant using Milvus

How Inkeep and Milvus Built a RAG-driven AI Assistant for Smarter Interaction

HNSWlib is an open-source C++ and Python library implementation of the HNSW algorithm, which is used for fast approximate nearest neighbor search. 

large_What is HNSWlib.png

medium_What is HNSWlib.png

small_What is HNSWlib.png

thumbnail_What is HNSWlib.png

What is HNSWlib.png

HNSWlib: A Graph-based Library for Fast ANN Search

Understanding HNSWlib: A Graph-based Library for Fast Approximate Nearest Neighbor Search

In this article, we explored various RAG components and discussed the approaches with optimal performance in each component. 

large_Searching for Best Practices in Retrieval-Augmented Generation.png

medium_Searching for Best Practices in Retrieval-Augmented Generation.png

small_Searching for Best Practices in Retrieval-Augmented Generation.png

thumbnail_Searching for Best Practices in Retrieval-Augmented Generation.png

Searching for Best Practices in Retrieval-Augmented Generation.png

Best Practices for Building RAG Apps

Best Practices in Implementing Retrieval-Augmented Generation (RAG) Applications 

In this article, we'll explore the evolution of MAS from a methodological or approach-based perspective.

large_The Evolution of Multi-Agent Systems_ From Early Neural Networks to Modern Distributed Learning ( Part II).png

medium_The Evolution of Multi-Agent Systems_ From Early Neural Networks to Modern Distributed Learning ( Part II).png

small_The Evolution of Multi-Agent Systems_ From Early Neural Networks to Modern Distributed Learning ( Part II).png

thumbnail_The Evolution of Multi-Agent Systems_ From Early Neural Networks to Modern Distributed Learning ( Part II).png

The Evolution of Multi-Agent Systems_ From Early Neural Networks to Modern Distributed Learning ( Part II).png

The Evolution of Multi-Agent Systems From a Methodological Perspective

The Evolution of Multi-Agent Systems: From Early Neural Networks to Modern Distributed Learning (Methodological)

In this article, we'll discuss the evolution of MAS from its early days to the most recent developments from an algorithmic perspective. 

large_The Evolution of Multi-Agent Systems_ From Early Neural Networks to Modern Distributed Learning ( Part I).png

medium_The Evolution of Multi-Agent Systems_ From Early Neural Networks to Modern Distributed Learning ( Part I).png

small_The Evolution of Multi-Agent Systems_ From Early Neural Networks to Modern Distributed Learning ( Part I).png

thumbnail_The Evolution of Multi-Agent Systems_ From Early Neural Networks to Modern Distributed Learning ( Part I).png

The Evolution of Multi-Agent Systems_ From Early Neural Networks to Modern Distributed Learning ( Part I).png

The Evolution of Multi-Agent Systems From an Algorithmic Perspective

The Evolution of Multi-Agent Systems: From Early Neural Networks to Modern Distributed Learning (Algorithmic) 

PagedAttention and vLLM solve important challenges in serving LLMs, particularly the high costs and inefficiencies in GPU memory usage when using it for inference. 


large_Efficient Memory Management for Large Language Model Serving with PagedAttention (1).png

medium_Efficient Memory Management for Large Language Model Serving with PagedAttention (1).png

small_Efficient Memory Management for Large Language Model Serving with PagedAttention (1).png

thumbnail_Efficient Memory Management for Large Language Model Serving with PagedAttention (1).png

Efficient Memory Management for Large Language Model Serving with PagedAttention (1).png

Efficient Memory Management for Large Language Model Serving with PagedAttention

Deep residual learning solves the degradation problem, allowing us to train a neural network while still potentially improving its performance.

large_Deep Residual Learning for Image Recognition.png

medium_Deep Residual Learning for Image Recognition.png

small_Deep Residual Learning for Image Recognition.png

thumbnail_Deep Residual Learning for Image Recognition.png

Deep Residual Learning for Image Recognition.png

What is Deep Residual Learning? 

Deep Residual Learning for Image Recognition

Zilliz Cloud Serverless allows users to store, index, and query massive amounts of vectors at only a fraction of the cost while keeping a competitive level of performance.  


large_Up to 50x Cost Savings for Building GenAI Apps Using Zilliz Cloud Serverless (1).png

medium_Up to 50x Cost Savings for Building GenAI Apps Using Zilliz Cloud Serverless (1).png

small_Up to 50x Cost Savings for Building GenAI Apps Using Zilliz Cloud Serverless (1).png

thumbnail_Up to 50x Cost Savings for Building GenAI Apps Using Zilliz Cloud Serverless (1).png

Up to 50x Cost Savings for Building GenAI Apps Using Zilliz Cloud Serverless (1).png

Up to 50x Cost Cut for Vector Search using Zilliz Cloud Serverless 

Up to 50x Cost Savings for Building GenAI Apps Using Zilliz Cloud Serverless

GPU-accelerated vector search through NVIDIA's cuVS library and CAGRA algorithm are highly beneficial for optimizing AI app performance in production. 

large_Milvus on GPUs with NVIDIA RAPIDS cuVS.png

medium_Milvus on GPUs with NVIDIA RAPIDS cuVS.png

small_Milvus on GPUs with NVIDIA RAPIDS cuVS.png

thumbnail_Milvus on GPUs with NVIDIA RAPIDS cuVS.png

Milvus on GPUs with NVIDIA RAPIDS cuVS.png

Supercharging Vector Search: Milvus on GPUs with NVIDIA RAPIDS cuVS

Milvus on GPUs with NVIDIA RAPIDS cuVS 

In this article, we'll explore time series databases in detail and walk you through a use case where we'll store time-series data in InfluxDB, query the data, transform it into vector embeddings, store the embeddings in Milvus, and finally perform a similarity search with Milvus. 

large_Improving Analytics with Time Series and Vector Databases (1).png

medium_Improving Analytics with Time Series and Vector Databases (1).png

small_Improving Analytics with Time Series and Vector Databases (1).png

thumbnail_Improving Analytics with Time Series and Vector Databases (1).png

Improving Analytics with Time Series and Vector Databases (1).png

We explore the combination of time series databases and vector databases, focusing on the use case in Internet of Things (IoT) applications.

Improving Analytics with Time Series and Vector Databases

In this article, we'll discuss a specific information retrieval method known as Boolean Retrieval Models. 

large_Aug 24 ——Understanding Boolean Retrieval Models in Information Retrieval.png

medium_Aug 24 ——Understanding Boolean Retrieval Models in Information Retrieval.png

small_Aug 24 ——Understanding Boolean Retrieval Models in Information Retrieval.png

thumbnail_Aug 24 ——Understanding Boolean Retrieval Models in Information Retrieval.png

Aug 24 ——Understanding Boolean Retrieval Models in Information Retrieval.png

Boolean retrieval models are an information retrieval method for retrieving relevant documents using a set of Boolean logic operators, eg. AND, OR, and NOT.


What are Boolean Retrieval Models in Information Retrieval 

Understanding Boolean Retrieval Models in Information Retrieval 

In this article, we will discuss a method, Constitutional AI (CAI), presented by the Anthropic team in their paper “Constitutional AI: Harmlessness from AI Feedback".


large_Aug 16  -- Constitutional AI_ Harmlessness from AI Feedback.png

medium_Aug 16  -- Constitutional AI_ Harmlessness from AI Feedback.png

small_Aug 16  -- Constitutional AI_ Harmlessness from AI Feedback.png

thumbnail_Aug 16  -- Constitutional AI_ Harmlessness from AI Feedback.png

Aug 16  -- Constitutional AI_ Harmlessness from AI Feedback.png

Constitutional AI (CAI) is a concept in artificial intelligence focused on ensuring that AI systems operate in alignment with ethical and legal principles


What is Constitutional AI (CAI)? 

Constitutional AI: Harmlessness from AI Feedback

Multimodal RAG extends RAG by accepting data from different modalities as context. Learn how to build one with Gemini 1.5, BGE-M3, Milvus, and LangChain. 


large_Aug 05 --Building a Multimodal RAG with  Gemini 1.5, BGE-M3, Milvus Lite, and LangChain.png

medium_Aug 05 --Building a Multimodal RAG with  Gemini 1.5, BGE-M3, Milvus Lite, and LangChain.png

small_Aug 05 --Building a Multimodal RAG with  Gemini 1.5, BGE-M3, Milvus Lite, and LangChain.png

thumbnail_Aug 05 --Building a Multimodal RAG with  Gemini 1.5, BGE-M3, Milvus Lite, and LangChain.png

Aug 05 --Building a Multimodal RAG with  Gemini 1.5, BGE-M3, Milvus Lite, and LangChain.png

Build a Multimodal RAG with Gemini, BGE-M3, Milvus and LangChain

Building a Multimodal RAG with Gemini 1.5, BGE-M3, Milvus Lite, and LangChain 

In this piece, we'll talk about a method called LLM-Eval, which is used to evaluate the response quality of an LLM. 

large_Aug 16  -- LLM-Eval_ A Streamlined Approach to Evaluating LLM Conversations.png

medium_Aug 16  -- LLM-Eval_ A Streamlined Approach to Evaluating LLM Conversations.png

small_Aug 16  -- LLM-Eval_ A Streamlined Approach to Evaluating LLM Conversations.png

thumbnail_Aug 16  -- LLM-Eval_ A Streamlined Approach to Evaluating LLM Conversations.png

Aug 16  -- LLM-Eval_ A Streamlined Approach to Evaluating LLM Conversations.png

LLM-Eval is an approach to simplifying and automating the evaluation of LLM conversation quality.


LLM-Eval: A Simplified Approach to Evaluating LLM Conversations 

LLM-Eval: A Streamlined Approach to Evaluating LLM Conversations 

Retrieval Augmented Generation (RAG) is a method for mitigating LLM hallucinations. Learn how to build a chatbot RAG with Milvus, Llama3, and LlamaIndex. 

large_Building RAG with Milvus Lite, Llama3, and LlamaIndex.png

medium_Building RAG with Milvus Lite, Llama3, and LlamaIndex.png

small_Building RAG with Milvus Lite, Llama3, and LlamaIndex.png

thumbnail_Building RAG with Milvus Lite, Llama3, and LlamaIndex.png

Building RAG with Milvus Lite, Llama3, and LlamaIndex.png

How to Build Retrieval Augmented Generation (RAG) with Milvus Lite, Llama3 and LlamaIndex 

Building RAG with Milvus Lite, Llama3, and LlamaIndex

Regularization prevents a machine-learning model from overfitting during the training process. We'll discuss its concept and key regularization techniques. 

large_Understanding Regularization in Neural Networks.png

medium_Understanding Regularization in Neural Networks.png

small_Understanding Regularization in Neural Networks.png

thumbnail_Understanding Regularization in Neural Networks.png

Understanding Regularization in Neural Networks.png

Understanding Neural Network Regularization and Key Regularization Techniques

Understanding Regularization in Neural Networks 

A step-by-step process to build an efficient and production-ready vector search pipeline using Databricks Spark and Milvus. 

large_Building Production Ready Search Pipelines with Spark and Milvus.png

medium_Building Production Ready Search Pipelines with Spark and Milvus.png

small_Building Production Ready Search Pipelines with Spark and Milvus.png

thumbnail_Building Production Ready Search Pipelines with Spark and Milvus.png

Building Production Ready Search Pipelines with Spark and Milvus.png

Building Production Ready Search Pipelines with Spark and Milvus

LangChain.rb eliminates the hassle of full-stack developers switching to another programming language when they want to leverage LLMs in their web applications.

large_July 09 ——Building an End-to-End GenAI App with Ruby  and Milvus.png

medium_July 09 ——Building an End-to-End GenAI App with Ruby  and Milvus.png

small_July 09 ——Building an End-to-End GenAI App with Ruby  and Milvus.png

thumbnail_July 09 ——Building an End-to-End GenAI App with Ruby  and Milvus.png

July 09 ——Building an End-to-End GenAI App with Ruby  and Milvus.png

Building an End-to-End GenAI App with Ruby and Milvus 

With Snowflake's Snowpark Container Service (SPCS), users can now run Milvus within the Snowflake ecosystem, allowing them to easily interact with Milvus using data stored in Snowflake.

large_July 02 ——Building RAG with Self-Deployed Milvus Vector Database and Snowpark Container Services.png

medium_July 02 ——Building RAG with Self-Deployed Milvus Vector Database and Snowpark Container Services.png

small_July 02 ——Building RAG with Self-Deployed Milvus Vector Database and Snowpark Container Services.png

thumbnail_July 02 ——Building RAG with Self-Deployed Milvus Vector Database and Snowpark Container Services.png

July 02 ——Building RAG with Self-Deployed Milvus Vector Database and Snowpark Container Services.png

Building RAG with Self-Deployed Milvus and Snowpark Container Services

Building RAG with Self-Deployed Milvus Vector Database and Snowpark Container Services

This article discusses several options for customizing LLMs to enhance their performance on specific tasks. 

large_June 14——Introduction to LLM Customization.png

medium_June 14——Introduction to LLM Customization.png

small_June 14——Introduction to LLM Customization.png

thumbnail_June 14——Introduction to LLM Customization.png

June 14——Introduction to LLM Customization.png

Introduction to LLM Customization

This article discussed how Delivery Hero used AI models to generate high-quality food images to improve user experience and conversion rate. 

large_How Delivery Hero Implemented the Safety System for AI-Generated Images.png

medium_How Delivery Hero Implemented the Safety System for AI-Generated Images.png

small_How Delivery Hero Implemented the Safety System for AI-Generated Images.png

thumbnail_How Delivery Hero Implemented the Safety System for AI-Generated Images.png

How Delivery Hero Implemented the Safety System for AI-Generated Images.png

How Delivery Hero Implemented the Safety System for AI-Generated Images

In this post, we'll explain how to extract content from a website and use it as context for LLMs in a RAG application. However, before doing so, we need to understand website fundamentals.

large_May 30--A Beginner's Guide to Website Chunking and Embedding for Your RAG Applications.png

medium_May 30--A Beginner's Guide to Website Chunking and Embedding for Your RAG Applications.png

small_May 30--A Beginner's Guide to Website Chunking and Embedding for Your RAG Applications.png

thumbnail_May 30--A Beginner's Guide to Website Chunking and Embedding for Your RAG Applications.png

May 30--A Beginner's Guide to Website Chunking and Embedding for Your RAG Applications.png

This post explains how to extract content from a website and use it as context for LLMs in a RAG application. However, before doing so, we need to understand website fundamentals.

A Beginner's Guide to Website Chunking and Embedding for Your RAG Applications

In this post, we'll explore several vector indexing strategies that can be used to efficiently perform similarity search, even in scenarios where we have large amounts of data and multiple constraints to consider.


large_How to Pick a Vector Index in Your Milvus Instance_ A Visual Guide.png

medium_How to Pick a Vector Index in Your Milvus Instance_ A Visual Guide.png

small_How to Pick a Vector Index in Your Milvus Instance_ A Visual Guide.png

thumbnail_How to Pick a Vector Index in Your Milvus Instance_ A Visual Guide.png

How to Pick a Vector Index in Your Milvus Instance_ A Visual Guide.png

How to Pick a Vector Index in Your Milvus Instance: A Visual Guide

In general, there are two types of vectors: dense vectors and sparse vectors. While they can be utilized for similar tasks, each has advantages and disadvantages. In this post, we will delve into two popular variants of sparse embedding: BM25 and SPLADE. 
