TensorFlow 分布式训练

技术分享 手册留言
10

我没有试过既多机又多卡的环境，不过看官方文档上的说明，似乎 MultiWorkerMirroredStrategy 就够了

tf.distribute.experimental.MultiWorkerMirroredStrategy is very similar to MirroredStrategy . It implements synchronous distributed training across multiple workers, each with potentially multiple GPUs.