Rohan Pandey

LlamaGym. Fine-tune LLM agents with online reinforcement learning

interrupting-cow. 🐮📢 The first AI voice assistant that interrupts *you*

sanskrit-ocr. Python

complexity-scaling. gzip Predicts Data-dependent Scaling Laws

MindMapResearch. Development of an NLP algorithm to temporally track patient emotional state using digital footprint

deredact. De-redacting Elon's Email with Character-count Constrained Llama2 Decoding

gpt2-explorer. gpt-2 latent representation explorer, runs in browser & visualizes similarity as graph connectedness

SATScale. Towards scalable solutions for boolean satisfiability

yowza. Automatic generation of listicles from AskReddit threads

i-am-a-strange-dataset. Repository for "I am a Strange Dataset: Metalinguistic Tests for Language Models"

ask-a-lecture. Search engine for video lectures that increases information retrieval efficiency by >90%

celery. Using AI and Big Data to optimize sustainability and profitability for small businesses

prime-rl. prime-rl is a codebase for decentralized RL training at scale

basehacks. HTML

atropos. Atropos is a Language Model Reinforcement Learning Environments framework for collecting and evaluating LLM trajectories through diverse environments

APCS_Work. All my AP CS work in Java from high school; deleting from local machine

nano-r1-speedrun. Single File, Single GPU, From Scratch, Efficient, Full Parameter Tuning library for "RL for LLMs"

BitFix. BitFix is a service that connects open source social good projects that need more hands on deck with volunteers that want to donate their time, skill, and energy to such projects

tinygrad. You like pytorch? You like micrograd? You love tinygrad! ❤️

Alleviair. Disaster Alleviation via Air

vidvas. Sanskrit text database and navigator

CACR. Cross-modal Attention Congruence Regularization implementation

FbIrl. Augmented Reality Experiment in Futuristic Social Media