1. Distilbert, a distilled version of bert: smaller, faster, cheaper and lighter;sanh,2019
2. Emerging Properties in Self-Supervised Vision Transformers
3. FTRANS
4. I-bert: Integer-only bert quantization;kim;ICML,2021
5. An fpga-based transformer accelerator using output block stationary dataflow for object recognition applications;zhao;IEEE TCAS-II Express Briefs,2022