We show for the first time that large-scale generative pretrained transformer (GPT) family
models can be pruned to at least 50% sparsity in one-shot, without any retraining, at minimal …

Diffusion models are powerful, but they require a lot of time and data to train. We propose
Patch Diffusion, a generic patch-wise training framework, to significantly reduce the training …

During image editing, existing deep generative models tend to re-synthesize the entire
output from scratch, including the unedited regions. This leads to a significant waste of …

High-resolution dense prediction enables many appealing real-world applications, such as
computational photography, autonomous driving, etc. However, the vast computational cost …

As the computing paradigm shifts from cloud computing to end-edge-cloud computing, it
also supports artificial intelligence evolving from a centralized manner to a distributed one …

We consider the problem of model compression for deep neural networks (DNNs) in the
challenging one-shot/post-training setting, in which we are given an accurate trained model …

On-device training enables the model to adapt to new data collected from the sensors by
fine-tuning a pre-trained model. Users can benefit from customized AI models without having …