Making machines think.

Hi, I'm Swarit. I'm a Computer Science & Engineering junior specializing in AI & ML. My primary focus is on Natural Language Processing (NLP) and reasoning systems. I am passionate about building real life usecases of AI.

Beyond AI, I like to watch cinema.

Selected Work

• GPT-2 Python · NLP

  An implementation of the 124M-parameter GPT-2 model built from first principles in PyTorch.

• Transformer Python · NLP

  An 11M-parameter decoder-only GPT-style language model built from first principles in PyTorch, using causal self-attention to generate character-level fantasy text.

• Foundational NLP Architectures Python · NLP

  A collection of character-level language models, including a Bigram model, the Neural Probabilistic Language Model (Bengio et al., 2003), and WaveNet for text generation, built entirely from scratch.

• Foundational ML Architectures Python · NumPy

  Core ML models—including Neural Networks, Linear/Logistic Regression, Softmax, and Anomaly Detection—built from the ground up to showcase the underlying calculus and linear algebra.

Specializations & Courses

• Deep Learning Specialization DeepLearning.AI

  Went deep into NLP, Convolutional Neural Networks, Sequence Models & Improving Neural Networks.

• Machine Learning Specialization Stanford University

  Covered Advanced Learning Algorithms, Supervised Machine Learning, and Unsupervised Learning & Recommenders.

• Mathematics for Machine Learning DeepLearning.AI

  Mastery in Calculus, Linear Algebra, Probability, and Statistics applied to Machine Learning.