Benchmarking Local LLM Performance on a MacBook

Measuring how quantization, serving engine, and prefix caching affect a coding-agent workload on a 24 GB M4 Mac — TTFT, throughput, and HumanEval pass@1 across vllm-metal and Ollama.

Text Generation & Word Prediction using RNN

There are a lot of beautiful things about neural networks and one of them is Recurrent Neural Networks (RNN). RNNs have been used in a variety of fields lately and have given very good results.

Deep Reinforcement Learning to play Flappy Bird using A3C algorithm

Overview

This project uses Asynchronous advantage actor-critic algorithm (A3C) to play Flappy Bird using Keras deep learning library. The details of this algorithm are mentioned in this paper by Google DeepMind. The code for this project can be found in this GitHub repository.