FromSystem

The origin of the Non-IID phenomenon is the personalization of users, who generate the Non-IID data. With Non-IID (Not Independent and Identically Distributed) issues existing in the federated learning setting, a myriad of approaches has been proposed to crack this hard nut. In contrast, the personalized federated learning may take the advantage…

PyTorch image models, scripts, pretrained weights -- ResNet, ResNeXT, EfficientNet, EfficientNetV2, NFNet, Vision Transformer, MixNet, MobileNet-V3/V2, RegNet, DPN, CSPNet, and more

MNN is a blazing fast, lightweight deep learning framework, battle-tested by business-critical use cases in Alibaba

[ICLR 2020] Once for All: Train One Network and Specialize it for Efficient Deployment

ConvNet training using pytorch

Low-precision matrix multiplication

Training MNIST with CoreML on Device

 Now we have become very big, Different from the original idea. Collect premium software in various categories.

RDMA core userspace libraries and daemons