1 为什么TensorFlow使用图的方式?
Why graphs?
1 Save computation(only run subgraphs that lead to the values you want to fetch)
节约计算量:原因在于图的回溯,当需要某节点值时,才会向前计算,通过session进行计算;
2 Break computation into small, differential pieces to facilitates auto-differentiation
将计算任务划分,加快求导:在深度神经网络中,使用梯度下降法更新权值和偏置;
3 Facilitate distributed computation, spread the work across multiple CPUs,GPUs, or devices
分布式计算的实现:如上,由于图的特性,将计算分成几块来并行运算,提高速度;
4 Many common machine learning models are commonly taught and visualized as directed graphs already
机器学习模型通常使用图形式进行描述;
Tensorflow的一些散知识点
最新推荐文章于 2024-07-06 02:15:17 发布