Active Projects

• FMEngine - Utilities to Train Very Large Language Models.

Datasets

• RedPajama - Large-Scale Pre-training Dataset for Language Modeling.
• Crosstalk Dataset.
• Covid-19 Twitter Dataset.
• Open Source Community Analytical Dataset.

Past (inactive) Projects

• Foodie: Comprehensive Food Recommendation.
• Tinynet: Implement Neural Network From Scratch.
• Covid-Sentiment: Analyse People’s Response to Covid-19 from Tweets. Demo
• Industry AI: Online Annotation for Image Datasets.
• AID - One Stop Machine Learning Model Management System.
• EurusDB.
• SHiFT - Search Engine for Machine Learning.
• Data-Centric vision Benchmark for Training Data Debugging.