Akshay πŸš€ Profile picture
Simplifying LLMs, MLOps, Python & Machine Learning for you! β€’ AI Engineering @LightningAI β€’ Lead Data Scientist β€’ BITS Pilani β€’ 3 Patents
Ella Sanders Profile picture Rodolpho Gurgel Profile picture Sujoy De Profile picture valuepiper Profile picture Shashi Kumar Profile picture 13 subscribed
Jul 26 β€’ 6 tweets β€’ 2 min read
Backpropagation in PyTorch, clearly explained: Backpropagation is the key to optimizing our neural network by adjusting it's weights.

And it's done by calculating gradients of the loss function w.r.t. these weights!

Today, we learn how to do this using PyTorch...πŸ‘‡ Image
Jul 24 β€’ 11 tweets β€’ 4 min read
Let's build a RAG app using MetaAI's Llama-3.1 (100% local): Before we begin, take a look at what we're about to create!

Here's what you'll learn:

- @Llama_Index for orchestration
- @qdrant_engine to self-host a vector DB
- @Ollama for locally serving Llama-3.1
- @LightningAI for development & hosting

Let's go! πŸš€
Jul 23 β€’ 11 tweets β€’ 3 min read
SQL Joins clearly explained: Let's setup 2 DataFrames to perform merge operations & name them:

β€’ left
β€’ right

A, B, C are the common keys βœ…

Check this πŸ‘‡ Image
Jul 23 β€’ 11 tweets β€’ 3 min read
Illustrated Guide to Tensor Parallelism (supercharge your LLM training): What is Tensor Parallelism?

Tensor Parallelism is form of model parallelism. It splits individual tensors across GPUs for efficient computation and memory use.

Perfect for training LLMs!

Let's simplify it today! πŸš€ Image
Jul 18 β€’ 9 tweets β€’ 3 min read
Let's implement & train this neural network step-by-step, from scratch using PyTorch!

1/n Image First of all let's define our model in PyTorch:

2/n Image
Jul 17 β€’ 11 tweets β€’ 4 min read
Let's build an advanced "Chat with your code" RAG application (100% local): Before we begin, take a look at what we're about to create!

Here's what you'll learn:

- @Llama_Index for orchestration
- @Qdrant_Engine to self-host a vectorDB
- LlamaIndex's advanced code parsers
- @Ollama for serving LLMs locally

Let's go! πŸš€
Jul 16 β€’ 10 tweets β€’ 3 min read
Multiprocessing in Python clearly explained: Ever felt like your Python code could run faster❓

Multiprocessing might be the solution you're looking for!

Today, I'll simplify it for you in this step-by-step guide.

Let's go! πŸš€ Image
Jul 15 β€’ 10 tweets β€’ 4 min read
Autoencoders are one of my favourite neural networks!

Today, I'll clearly explain:

- What they are❓
- And how they work❓

Let's go! πŸš€

1/n Image Autoencoders have two main parts:

1️⃣ Encoder: Compresses the input into a dense representation (latent space)

2️⃣ Decoder: Reconstructs the input from this dense representation.

The idea is to make the reconstructed output as close to the original input as possible:πŸ‘‡

2/n Image
Jul 12 β€’ 7 tweets β€’ 3 min read
ML models can only be as good as the data they're trained on!

Introducing Datalab, now you can automatically detect:

- outliers
- label errors
- (near) duplicates
- low-quality/non-IID sampling

Developed at MIT, Datalab works with all types of data & any trained model!

1/n Image How to use Datalab❓

Datalab works with any ML model you have already trained!

It's like a magic wand! πŸͺ„

Inspecting your dataset with Datalab merely requires the code below! πŸ‘‡

2/n Image
Jul 11 β€’ 11 tweets β€’ 4 min read
Let's build a "Chat with your docs" RAG application, step-by-step: Before we begin, take a look at what we're about to create!

We'll be using:

- @Cohere highly capable ⌘R+ as the LLM
- @Llama_Index for orchestration
- @Streamlit for the UI

Everything in just ~170 lines of Python code, that I've shared at the end! πŸ”₯

Let's go! πŸš€
Jul 9 β€’ 7 tweets β€’ 3 min read
5 GitHub repositories that will give you superpowers as an AI/ML Engineer: 1️⃣ Awesome Artificial Intelligence

A curated list of Artificial Intelligence:

- courses
- books
- video lectures
- and papers with code

Check this out πŸ‘‡
github.com/owainlewis/awe…
Jul 8 β€’ 9 tweets β€’ 3 min read
This is the future of building RAGs. Let me introduce you to DSPy today: In terms of analogy, DSPy is to RAG as PyTorch is to DNNs.

DSPy : RAG :: PyTorch : DNNs

To understand DSPy, we need to grasp three key concepts:

- Signatures
- Modules
- Optimizers

Let's delve into each one! πŸš€ Image
Jul 4 β€’ 9 tweets β€’ 3 min read
Let's make RAG 40x faster and 32x memory efficient: To achieve this we'll be leveraging binary quantization! πŸš€

And here's what you'll learn, today:

- Intuitive explanation of Binary Quantization (BQ)
- Self-host @qdrant_engine with BQ enabled
- Search over 36M+ vectors in <50ms πŸ”₯

Let's go! πŸš€
Jul 3 β€’ 7 tweets β€’ 3 min read
Tensors in PyTorch, clearly explained: Tensors are the fundamental building blocks for performing mathematical operations in deep learning models.

Today, I will provide a comprehensive explanation with illustrative code examples.

Let's go! πŸš€ Image
Jul 2 β€’ 10 tweets β€’ 3 min read
7 projects that every AI engineer must explore: 1️⃣ Chat with your code using RAG

A step by step guide to building a RAG application using @llama_index!

Check this outπŸ‘‡
lightning.ai/lightning-ai/s…
Jun 27 β€’ 7 tweets β€’ 3 min read
Eigenvalues & Eigenvectors clearly explained: The concept of eigenvalues & eigenvectors is widely known yet poorly understood!

Today, I'll clearly explain their meaning & significance.

Let's go! πŸš€ Image
Jun 24 β€’ 10 tweets β€’ 3 min read
f-strings in Python clearly explained: f-strings were introduced in Python 3.6 and have since become a favorite among developers for their simplicity and readability.

Today, we'll start with the basics and dive into all the ninja tricks of using f-strings.

Let's go! πŸš€ Image
Jun 20 β€’ 10 tweets β€’ 3 min read
Multithreading in Python clearly explained: Ever felt like your Python code could run faster❓

Multithreading might be the solution you're looking for!

Today, I'll simplify it for you in this step-by-step guide.

Let's go! πŸš€ Image
Jun 17 β€’ 11 tweets β€’ 3 min read
Object oriented programming in Python, clearly explained: We break it down to 6 important concepts:

- Object 🚘
- Class πŸ—οΈ
- Inheritance 🧬
- Encapsulation πŸ”
- Abstraction 🎭
- Polymorphism πŸŒ€

Let's take them one-by-one... πŸš€ Image
Jun 15 β€’ 12 tweets β€’ 4 min read
I started my career in Data Science back in 2016 ⏳

Here's a detailed roadmap for those starting out today!

What's covered:
- Python
- Machine Learning
- Maths for ML
- ML Books
- MLOps
- LLMs/AI Engineering

Read more...πŸ‘‡ Image 1️⃣ Python

If you are new to programming and just getting started.

There isn't a better place to learn Python than David J Malan's CS50p.

Beautiful explanations and great projects.
It's a complete package ⚑️

Check this out πŸ‘‡
edx.org/course/cs50s-i…
Jun 13 β€’ 9 tweets β€’ 3 min read
Let's learn how to evaluate a RAG application: Here's what we'll do today:

- Build a RAG pipeline using @llama_index
- Evaluate it with @ragas_io
- Implement observability using @ArizePhoenix

Before we dive in, check out this demo: