Fartash Faghri is a machine learning researcher at Apple Machine Learning Research (MLR). His research centers on enhancing the efficiency of foundation model training. Specifically, he has contributed to improving quality and effectiveness of training datasets, developing techniques for training on-device foundation models, and designing benchmarks and novel strategies for large-scale continual learning. Fartash joined Apple in 2021 following the completion of his PhD at the University of Toronto. His doctoral research delved into the development of fast and robust training methods.

News

• 2025/06/24: TiC-LM is selected for Oral presentation (Top 8%) at ACL 2025 in Vienna, Austria.
• 2025/06/14: FastVLM will be presented at CVPR 2025.
• 2024/12/13: DataComp-LM will be presented at NeurIPS 2024.
• 2024/06/20: MobileCLIP will be presented at CVPR 2024.
• 2024/05/10: TiC-CLIP will be presented at ICLR 2024.