Sachin Mehta

KV Prediction for Improved Time to First Token

arXiv 2024

Bytes Are All You Need: Transformers Operating Directly On File Bytes

arXiv 2023

TiC-CLIP: Continual Training of CLIP Models

arXiv 2023

Reinforce Data, Multiply Impact: Improved Model Accuracy and Robustness with Dataset Reinforcement

ICCV 2023 1

MedICaT: A Dataset of Medical Images, Captions, and Textual References

Findings of the Association for Computational Linguistics 2020

DeLighT: Deep and Light-weight Transformer

delight-deep-and-light-weight-transformer

Mohammad Rastegari

Ali Farhadi

CEO

Fartash Faghri

Hadi Pouransari

Hannaneh Hajishirzi

professor

Maxwell Horton

Mehrdad Farajtabar

Oncel Tuzel

Ben Bogin

Chenfan Sun