all AI news
Q-ViT: Fully Differentiable Quantization for Vision Transformer. (arXiv:2201.07703v1 [cs.CV])
Jan. 20, 2022, 2:10 a.m. | Zhexin Li, Tong Yang, Peisong Wang, Jian Cheng
cs.CV updates on arXiv.org arxiv.org
In this paper, we propose a fully differentiable quantization method for
vision transformer (ViT) named as Q-ViT, in which both of the quantization
scales and bit-widths are learnable parameters. Specifically, based on our
observation that heads in ViT display different quantization robustness, we
leverage head-wise bit-width to squeeze the size of Q-ViT while preserving
performance. In addition, we propose a novel technique named switchable scale
to resolve the convergence problem in the joint training of quantization scales
and bit-widths. In …
More from arxiv.org / cs.CV updates on arXiv.org
Jobs in AI, ML, Big Data
Senior Machine Learning Engineer (MLOps)
@ Promaton | Remote, Europe
Research Associate (Data Science/Information Engineering/Applied Mathematics/Information Technology)
@ Nanyang Technological University | NTU Main Campus, Singapore
Associate Director of Data Science and Analytics
@ Penn State University | Penn State University Park
Student Worker- Data Scientist
@ TransUnion | Israel - Tel Aviv
Vice President - Customer Segment Analytics Data Science Lead
@ JPMorgan Chase & Co. | Bengaluru, Karnataka, India
Middle/Senior Data Engineer
@ Devexperts | Sofia, Bulgaria