🚀 Deep Learning Journey: Fast.ai

This repository contains my notes, homework, and projects from the Fast.ai Deep Learning for Coders course.

The goal of this repo is to track my progress, experiment with the fastai library, and implement state-of-the-art deep learning models.

📂 Repository Structure

The coursework is divided by lessons, with a mix of replicated course notebooks and my own independent experiments.

Lesson	Topic	My Notebook	Project/Experiment
01	Image Classification	Link	Is it a Bird? Classifier
02	Deployment & Production	Link	Bear Classifier App (HuggingFace)
03	Neural Net Foundations	Link	SGD from scratch
04	NLP Introduction	Link	IMDB Sentiment Analysis
...	...	...	...

🌟 Project Highlights

1. The Bear Classifier (Lesson 2)

A computer vision model capable of distinguishing between Grizzly, Black, and Teddy bears.

• Accuracy: 98%
• Data: Scraped using DuckDuckGo Search API.
• Live Demo: Click Here to Try it on HuggingFace Spaces
• Key Learning: How to clean data using the ImageClassifierCleaner and deploy a .pkl model.

2. Multi-Label Classification (Lesson 6)

[Insert Screenshot of results here]

• Goal: Identifying multiple attributes in a single satellite image.
• Architecture: ResNet50
• Key Learning: Understanding metrics for multi-label problems (F1 Score vs. Accuracy).