标签: tensorflow
当我阅读关于" Variable Distribution and Gradient Aggregation"对于张量流的高性能模型,我无法区分Parameter Server Variables和Replicated Variables in Distributed Training之间的区别。有人能解释清楚吗?