Meta AI

Software Engineer Fellow

2023/01 - 2023/04

Responsibilities & Achievements

• Integrated transformer architecture algorithms into Meta's AI open-source library (fairseq2), utilizing Python and Pytorch, based on AI literature. The library has been instrumental in state-of-the-art translation models and empowering researchers with the latest advancements to train Large Language Models.
• Refactored and modularized critical components of the transformer architecture such as multi-head attention.
• Researched and implemented algorithms for audio data augmentation, enriching the AI training process.