Selva Prabhakaran

Selva is an experienced Data Scientist and leader, specializing in executing AI projects for large companies. Selva started machinelearningplus to make Data Science / ML / AI accessible to everyone. The website enjoys 4 Million+ readership. His courses, lessons, and videos are loved by hundreds of thousands of students and practitioners.

153

Articles

Topics

Years Active

machine learning + How to Evaluate an LLM: Benchmarks, Metrics, and Practical Workflows machinelearningplus.com

25 min

NLP

How to Evaluate an LLM: Benchmarks, Metrics, and Practical Workflows

Evaluate your LLM using MMLU, MT-Bench, LLM-as-judge, and ROUGE. Covers lm-evaluation-harness, fine-tuned model comparison, and evaluation pitfalls. With code.

evaluate LLM LLM evaluation LLM-as-judge

machine learning + Unsloth Fine-Tuning — Train LLMs 2x Faster with 70% Less Memory machinelearningplus.com

20 min

Gen AI

Unsloth Fine-Tuning — Train LLMs 2x Faster with 70% Less Memory

Fine-tune LLMs with Unsloth — 2x faster training, 70% less GPU memory. Step-by-step guide with LoRA, QLoRA, code examples, and deployment tips.

fine-tune Llama LLM fine-tuning LoRA

machine learning + How to Evaluate LLMs — Metrics, Benchmarks & Python Code machinelearningplus.com

39 min

Gen AI

How to Evaluate LLMs — Metrics, Benchmarks & Python Code

Learn LLM evaluation from scratch -- benchmarks, metrics (BLEU, ROUGE, perplexity), LLM-as-judge, and custom pipelines with runnable Python code.

BLEU score Evaluation Metrics LLM benchmarks

machine learning + Fine-Tuning LLMs with LoRA and QLoRA in Python — A Complete Guide machinelearningplus.com

27 min

Deep Learning

Fine-Tuning LLMs with LoRA and QLoRA in Python — A Complete Guide

Fine-tune LLMs with LoRA and QLoRA in Python. Complete guide covering memory math, PEFT setup, 4-bit QLoRA, adapter merging, and common mistakes — with...

bitsandbytes fine-tuning LLMs LoRA

machine learning + DPO (Direct Preference Optimization) — A Simpler Alternative to RLHF machinelearningplus.com

40 min

Gen AI

DPO (Direct Preference Optimization) — A Simpler Alternative to RLHF

Align LLMs with human preferences using one loss function -- no reward model, no RL. Complete guide with derivation, PyTorch code, and DPO variants.

Direct Preference Optimization DPO fine-tuning

machine learning + How to Build a Custom Instruction Dataset for LLM Fine-Tuning machinelearningplus.com

28 min

Machine Learning

How to Build a Custom Instruction Dataset for LLM Fine-Tuning

Learn how to build a custom instruction dataset for LLM fine-tuning — covering Alpaca, ShareGPT, and DPO formats, quality filtering, synthetic data generation, token...

custom instruction dataset HuggingFace datasets instruction dataset fine-tuning

machine learning + How to Fine-Tune LLMs with LoRA in Python — A Complete Guide machinelearningplus.com

30 min

Deep Learning

How to Fine-Tune LLMs with LoRA in Python — A Complete Guide

Learn how to fine-tune large language models with LoRA in Python using PEFT and TRL — covers LoraConfig, QLoRA, SFTTrainer, model merging, and common...

fine-tune LLMs LLM training LoRA fine-tuning

machine learning + How LLMs Work in Python — Transformer Architecture Explained machinelearningplus.com

28 min

Deep Learning

How LLMs Work in Python — Transformer Architecture Explained

Learn how LLMs work from tokenization to text generation. Build self-attention from scratch in Python with verified code and 3 interactive exercises.

GPT Python tutorial how LLMs work Large Language Models

machine learning + Build a Python AI Chatbot with Memory — OpenAI Tutorial machinelearningplus.com

28 min

Python

Build a Python AI Chatbot with Memory — OpenAI Tutorial

Learn how to build a Python AI chatbot with memory using the OpenAI API. Covers conversation history, token management, streaming, and a complete reusable...

chatbot with memory conversation history python openai api tutorial

machine learning + Build Your First AI App with Python — Step-by-Step Tutorial machinelearningplus.com

25 min

Python

Build Your First AI App with Python — Step-by-Step Tutorial

Build your first AI app with Python and the OpenAI API. Step-by-step tutorial covering chat completions, streaming, error handling, and cost control — with...

build AI app python chat completions python gpt api python

Build a Custom Scikit-Learn Regression Model: Step-by-Step Guide

11 min

Machine Learning

Build a Custom Scikit-Learn Regression Model: Step-by-Step Guide

Creating custom regressors in scikit-learn means building your own machine learning models that follow scikit-learn’s API conventions, allowing them to work seamlessly with pipelines,...

machine learning + Understanding Confidence Intervals: A spelled out guide to clarify misconceptions machinelearningplus.com

4 min

Statistics

Understanding Confidence Intervals: A spelled out guide to clarify misconceptions

If you’ve ever read a scientific study, survey results, or even a political poll, you’ve probably encountered confidence intervals (CIs). They’re one of the...

Optimizing RAG Chunk Size: Your Definitive Guide to Better Retrieval Accuracy

14 min

Gen AI

Optimizing RAG Chunk Size: Your Definitive Guide to Better Retrieval Accuracy

Optimal chunk size for RAG systems typically ranges from 128-512 tokens, with smaller chunks (128-256 tokens) excelling at precise fact-based queries while larger chunks...

machine learning + Ridge Regression as MAP Estimation – Supporting notes machinelearningplus.com

5 min

Statistics

Ridge Regression as MAP Estimation – Supporting notes

This is one of the most beautiful connections in machine learning – let me break down exactly why Ridge regression is MAP estimation in...

Maximum A Posteriori (MAP) Estimation – Clearly Explained

14 min

Statistics

Maximum A Posteriori (MAP) Estimation – Clearly Explained

Maximum A Posteriori (MAP) estimation is a Bayesian method for finding the most likely parameter values given observed data and prior knowledge. Unlike maximum...

Relevant Segment Extraction (RSE) – Building better Context by assembling contiguous chunks for better RAG Performance

28 min

Gen AI

Relevant Segment Extraction (RSE) – Building better Context by assembling contiguous chunks for better RAG Performance

Relevant Segment Extraction (RSE) is a query-time post-processing technique that intelligently combines related text chunks into longer, coherent segments, providing LLMs with better context...

RAG

machine learning + Ollama Tutorial: Your Guide to running LLMs Locally machinelearningplus.com

5 min

Gen AI

Ollama Tutorial: Your Guide to running LLMs Locally

Ollama is a tool used to run the open-weights large language models locally. It’s quick to install, pull the LLM models and start prompting...

Large Language Models Ollama

machine learning + CUDA Programming: Do Large scale parallel computing in GPU from Scratch machinelearningplus.com

21 min

Deployment

CUDA Programming: Do Large scale parallel computing in GPU from Scratch

While GPUs ((Graphics Processing Unit) are in high demand in video games, with the rise of Large Language Models (LLMs), GPUs are in high...

CUDA

machine learning + Mutual information vs Cross Entropy machinelearningplus.com

4 min

Machine Learning

Mutual information vs Cross Entropy

Cross-entropy is a measure of error, while mutual information measures the shared information between two variable. Both concepts used in information theory, but they...

mutual information

machine learning + F Statistic Formula – Explained machinelearningplus.com

6 min

Statistics

F Statistic Formula – Explained

The F statistic is used in statistical hypothesis testing to determine if there are significant differences between group means. It is most commonly used...

Latest Articles

NLP
How to Evaluate an LLM: Benchmarks, Metrics, and Practical Workflows

Mar 7, 2026
Gen AI
Unsloth Fine-Tuning — Train LLMs 2x Faster with 70% Less Memory

Mar 7, 2026
Gen AI
How to Evaluate LLMs — Metrics, Benchmarks & Python Code

Mar 7, 2026

Selva Prabhakaran

How to Evaluate an LLM: Benchmarks, Metrics, and Practical Workflows

Unsloth Fine-Tuning — Train LLMs 2x Faster with 70% Less Memory

How to Evaluate LLMs — Metrics, Benchmarks & Python Code

Fine-Tuning LLMs with LoRA and QLoRA in Python — A Complete Guide

DPO (Direct Preference Optimization) — A Simpler Alternative to RLHF

How to Build a Custom Instruction Dataset for LLM Fine-Tuning

How to Fine-Tune LLMs with LoRA in Python — A Complete Guide

How LLMs Work in Python — Transformer Architecture Explained

Build a Python AI Chatbot with Memory — OpenAI Tutorial

Build Your First AI App with Python — Step-by-Step Tutorial

Build a Custom Scikit-Learn Regression Model: Step-by-Step Guide

Understanding Confidence Intervals: A spelled out guide to clarify misconceptions

Optimizing RAG Chunk Size: Your Definitive Guide to Better Retrieval Accuracy

Ridge Regression as MAP Estimation – Supporting notes

Maximum A Posteriori (MAP) Estimation – Clearly Explained

Relevant Segment Extraction (RSE) – Building better Context by assembling contiguous chunks for better RAG Performance

Ollama Tutorial: Your Guide to running LLMs Locally

CUDA Programming: Do Large scale parallel computing in GPU from Scratch

Mutual information vs Cross Entropy

F Statistic Formula – Explained

How to Evaluate an LLM: Benchmarks, Metrics, and Practical Workflows

Unsloth Fine-Tuning — Train LLMs 2x Faster with 70% Less Memory

How to Evaluate LLMs — Metrics, Benchmarks & Python Code

Machine Learning A-Z™: Hands-On Python & R In Data Science

Free Sample Videos:

Selva Prabhakaran

Latest Articles

Python.SQL. NumPy. All free.

Machine Learning A-Z™: Hands-On Python & R In Data Science

Free Sample Videos:

Machine Learning A-Z™: Hands-On Python & R In Data Science

Machine Learning A-Z™: Hands-On Python & R In Data Science

Machine Learning A-Z™: Hands-On Python & R In Data Science

Machine Learning A-Z™: Hands-On Python & R In Data Science

Machine Learning A-Z™: Hands-On Python & R In Data Science

Python.
SQL. NumPy.
All free.