1.NVIDIA / Megatron-LMOngoing research training transformer models at scalelarge-language-modelsmodel-paratransformersPythonOther license18K total4.7K forks+1,033 starsfrom 17,303 · score 6.94
2.microsoft / CNTKArchivedMicrosoft Cognitive Toolkit (CNTK), an open source deep-learning toolkitcognitive-toolkitcntkdeep-learningmachine-learningdeep-neural-networksneural-network+5 moreC++Other license18K total4.2K forks+26 starsfrom 17,557 · score 3.30
3.horovod / horovodArchivedDistributed training framework for TensorFlow, Keras, PyTorch, and Apache MXNet.tensorflowubermachine-learningmachinelearningmpibaidu+7 morePythonOther license15K total2.2K forks+7 starsfrom 14,667 · score 2.08
4.facebookresearch / PyTorch-BigGraphArchivedGenerate embeddings from large-scale graph-structured data.PythonOther license3.5K total453 forks+3 starsfrom 3,451 · score 1.39
5.tqchen / xgboosthttps://github.com/dmlc/xgboostC++Other license583 total255 forks+3 starsfrom 580 · score 1.39