Accuracy Boosters: Epoch-Driven Mixed-Mantissa Block Floating-Point for DNN Training

SB Harma, A Chakraborty, B Falsafi, M Jaggi… - arXiv preprint arXiv …, 2022 - arxiv.org
The unprecedented growth in DNN model complexity, size, and amount of training data has
led to a commensurate increase in demand for computing and a search for minimal …
被引用次数：6 相关文章

Accuracy Boosters: Epoch-Driven Mixed-Mantissa Block Floating-Point for DNN Training

SB Harma, C Sönmez, B Falsafi, M Jaggi, Y Oh - openreview.net
The unprecedented growth in DNN model complexity, size and the amount of training data
have led to a commensurate increase in demand for computing and a search for minimal …

Accuracy Boosters: Epoch-Driven Mixed-Mantissa Block Floating Point for DNN Training

SB Harma, A Chakraborty, B Falsafi, M Jaggi… - Architecture and System … - openreview.net
The unprecedented growth in DNN model complexity, size, and amount of training data has
led to a commensurate increase in demand for computing and a search for minimal …