# My Journey into AI

Hi, I’m Roshan Acharya, a Computer Engineering student from Nepal working on NLP and LLM-related projects and learning AI. I started this journey of learning AI exactly 1 year ago, a decision that changed how I learn, think and build.

Like many students, I was curious about AI. Terms like machine learning, deep learning, NLP and LLMs sounded exciting but I have no idea how it will go. I didn’t have a perfect roadmap or expert guidance.

## How It All Started

My journey began with a subject called Foundations of Data Science, which was part of my academic curriculum. Through this course, I was introduced to basic **Python, statistics, and classical machine learning** concepts, along with essential practices like data cleaning and **exploratory data analysis (EDA)**.

Instead of rushing to apply models, I focused on understanding why a model works, not just how to use it. This approach helped me build a strong conceptual foundation and develop a problem-solving mindset.

## 111Days of Challenge:

As I strengthened my basics, I participated 111Days of Challenge organized by **Code for Change**. This was the turning point where the habit of learning consistently was developed. I learned basic ML and DL concepts along with some hands on project like **Car Price Prediction, Laptop Price Prediction**, **Image Compressor**. These projects helped bridge the gap between theory and practice.

## Learning the Imoirtance of Data

One major realization during this phase was that **data collection** is one of the most important steps in any ML or AI project. I started exploring **web scraping**, collecting real-world data, cleaning it, and transforming it into something meaningful.

Working with messy, real data gave me confidence and prepared me to take the next step.

## Deep Learning and NLP

As I progressed in Machine Learning and Deep Learning, I became increasingly interested in how machines understand language, not just numbers and images. This curiosity led me to **Natural Language Processing (NLP)**. Along the way, I explored transformer based model. I am currently working on finetuning projects and RAG system.

## Still Learning, Still Training

I see this journey not as a destination.

I see it as a continuous learning process of learning, experimenting, and improving.

As long as the **loss keeps decreasing** and the **learning never stops**, I know the model is on the right path.

**Training in progress. 🚀**