Passionate about Machine Learning Research I build and re-implement deep learning systems from scratch to understand them deeply. Currently focusing on Multimodal Learning and contrastive representation learning.
- CLIP from Scratch: Re-implementing OpenAI's CLIP with custom ViT and Text Transformer backbones for multimodal contrastive learning.
- LinkedIn: Yasser-BOUHAI
- Email: yasserbouhai@gmail.com
"The goal is not to study AI, but to build things that force you to understand it."