Transformer-based pipeline for cell type classification across multimodal single-cell and spatial omics datasets. I owned model adaptation and experiment execution workflows during this internship research project.
- Extended a foundation transformer model pre-trained on 500K+ cells for supervised classification.
- Built configurable training, fine-tuning, and inference scripts across model sizes and datasets.
- Automated reproducible GPU experiment runs with Bash job scripts on remote compute instances.
Code is private due to research constraints associated with this project.