Simple API: Quickly build and train neural network models with a few lines of code. Customizable Architecture: Easily configure the number of layers, neurons per layer, activation functions, learning ...
A performance model for predicting the training time of ML (DLRM, CV, NLP, etc) models on single-GPU and multi-GPU platforms. This is also the code repo for the following two papers: (TPDS 2024) ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results