This repository contains various experiments inspired by curiosity and learning. Each experiment is organized into its own folder with a brief description below.
- Explores chunking data, generating embeddings, and performing retrieval tasks. Includes scripts and a Jupyter notebook for experimentation.
- Implements knowledge distillation on the MNIST dataset to transfer knowledge from a large model to a smaller one.
- Investigates concepts of subliminal learning and entanglement through theoretical and experimental approaches.(Baulabs study replication)
- Analyzes the Adam optimizer's behavior and performance in machine learning tasks.
- Measures how document size impacts retrieval time in MongoDB.
- Explores how various Python web frameworks process HTTP routes, focusing on sequential evaluation versus radix/tree-based compilation. Includes experiments with dynamic and static route handling and how in some the order of routes impacts the requests.