Delicious Blog

Creating different projects using Jupyter notebook

Building LLM part2-Processing Text

In this blog, we will go through chapter 2 of “Build a Large Language Model From Scratch” by Sebastian Raschka. This chapter is about working with text. It goes over…

Building LLM part1-Intro_to_LLMs

I am reading “Build a Large Language Model from scratch” by Sebastian Raschka in a cluster of stars study group. In this blog, I am writing about the first chapter, which…

A Student’s Perspective on Computational Biology

Computational biology is a very exciting field. As a student in college, I am still figuring out what it is and what it is capable of. This field moves quickly with an…

How to solve it

Over the winter break, I learned about Georgy Polya’s “How to solve it” through a course. “How to solve it” is a book about problem solving. As a mathematician, Polya solved…

Resnet Part2

We learned about what Resnet is in Resnet Blog. However, we didn’t train it, so we have no idea how much better it is than just convolutional layers. In this blog, we will…

MiniAI Utilities

I have been taking Part 2 of FastAI courses, which covers matrix multiplication to stable diffusion. In the course, we built MiniAI from scratch. MiniAI is a small and…

Resnet

In this blog, we will talk about Residual network (Resnet). Resnet came from Deep Residual Learning for Image Recognition by Kaiming He et al. We have seen Kaiming/He…

Scheduler

In this blog, we will learn about Pytorch optimizers and learning rate schedulers. From all the learning rate schedulers, we will learn about CosineAnnealingLR and OneCycleLR…

Optimizer

We use optimizers to update parameters. We will explore different strategies, such as SGD (Stochastic gradient descent), momentum, RMSProp, and Adam. This blog is based on…

Initialization part 2

This blog will teach us about Layer-wise Sequential Unit-Variance (LSUV), layer normalization, and batch normalization. This blog is based on lesson 17 of the FastAI course.

Initialization part 1

In this blog, let’s explore some initializing methods on weights to train better. Our model can train well with a good initialization because close to all the activations…

Pytorch Hooks

Today, we will go over Pytorch hooks. Hooks are callbacks, just functions that are called at a specific time. This blog is based on Practical Deep Learning lesson 16 and the…

Convolutional Neural Networks

Let’s learn about convolutional neural networks (CNN). Rather than using linear layers, we mostly use CNN on images. Compared to using linear layers, we can get more…

Learner Pt.2

In part 1, we looked at a basic learner and a callback learner. The basic learner’s code was hand-written without any flexibility. We would have to modify the learner if we…

Learner Pt.1

We learned about callbacks last time. In this blog, we will learn about a learner. It’s common to use a fit function to train a model. However, as we want to modify…

Callback

There are many callbacks used in deep learning. So, what is a callback? Simply, a callback is a function.

Hugging Face Features

In lesson 15 of the Practical Deep Learning For Coders, we used Hugging Face Datasets to download Fasion MNIST data and trained our model. I faced a problem here because I…

Fruit Multi-Classifier pt.2 Deployment

In this blog, we will deploy our model into Hugging Face Space using gradio. I have written a detailed blog here on how to deploy. This blog omits some steps.

Fruit Multi-Classifier pt.1 Training

We will train a fruit classifier using multi-label classification. It is a part of further research part of fastbook chapter6. Instead of creating a bear classifier, we will…

MNIST Neural Nets

In this blog, we will use neural networks to train MNIST dataset. We start with using pytorch and integrate fastai in the end.

MNIST base

In this blog, we will attempt to predict MNIST digits with very simple approach. Instead of using neural networks, we will find a mean of target images and compare an image…

MNIST in FastAI

Here, we will predict MNIST digits using FastAI. Then, we will go deeper and figure out what’s happening behind the scenes.

Saving Jupyter configuration on Paperspace.

Today, I will go over how to save Jupyter configurations on Paperspace so that we can keep using the customized shortcuts we saved.

Intro to Paddy competition

We will get started with Paddy competition from Kaggle.

Live coding 7

In this blog, I will cover how to setup kaggle on Paperspace.

Live coding 6

In this blog, I will cover how to setup mamba and bash_history in Paperspace. We also will go over a bit of fastbook and some terminal commands.

Live coding 5

In this blog, I will cover vim and ctags based on the Youtube video.

Live coding 4

In this blog, I will cover tips on Jupyter notebook and python debugger.

Live coding 3

This is a blog based on Jeremy Howard’s Youtube video on Live coding 3. Here, we will learn about $PATH, how conda/mamba works, how to setup pip and git in Paperspace.

Live coding 2

This is a blog based on Jeremy Howard’s Youtube video on Live coding 2. Here, we will learn about Git, tmux, and Jupyter notebook.

Live coding 1

This is a blog based on Jeremy Howard’s Youtube video on Live coding 1. Here, we will learn how to use terminal, install python the right way, and install packages using…

Alien vs. Ghost Pt.2 Deploying

In this article, we will deploy the model that we trained using FastAI on part 1 on Hugging Face Spaces with Gradio and nbdev.

Alien vs. Ghost Pt.1 Training

In this article, we will make a image classification model that attempts to differentiate between aliens and ghosts.