843 citations · 845 across the 3 of their papers we have counts for
4 papers
TorchTPU - A Hybrid Eager/Compiled Architecture for Native PyTorch Execution on TPU
Claudio Basile, Michael Voznesensky, Gunhyun Park +50
TorchTPU is an open-source collaboration between Google and Meta that introduces a native and principled PyTorch backend for Google TPUs. It enables seamless out-of-the-box model e…
Masterful: A Training Platform for Computer Vision Models
Samuel Wookey, Yaoshiang Ho, Tom Rikert +8
Masterful is a software platform to train deep learning computer vision models. Data and model architecture are inputs to the platform, and the output is a trained model. The platf…
The Human Visual System and Adversarial AI
Yaoshiang Ho, Samuel Wookey
This paper applies theories about the Human Visual System to make Adversarial AI more effective. To date, Adversarial AI has modeled perceptual distances between clean and adversar…
The Real-World-Weight Cross-Entropy Loss Function: Modeling the Costs of Mislabeling
Yaoshiang Ho, Samuel Wookey
In this paper, we propose a new metric to measure goodness-of-fit for classifiers, the Real World Cost function. This metric factors in information about a real world problem, such…