Forecasting Bitcoin returns through time series analysis, emphasizing sentiment analysis on news (using BERT LLM), social media, and Google search trends, with the final model based on Random Forest, augmented with engineered memory features.

MLS implemented for use on the Flamingo UAV swarm

Train OpenFlamingo vision-language models on LAION + MMC4 with distributed DDP/FSDP

ETL pipeline for converting raw LAION and MMC4 datasets into WebDataset tar shard format for OpenFlamingo training

No description available

No description available

This project explores big data, top organizations leveraging big data, and big data future in the age of LLM. Additionally, it delves into practical application using catch the pink flamingo dataset such as EDA, ML, and graph analytics using Neo4J.