Low-level API of TensorFlow includes tensor operation, graph, automatic differentiate, etc.
If we compare a model to a house, then these low-level APIs are the bricks.
We may use TensorFlow as the enhanced numpy through these low-level APIs.
TensorFlow provides a more complete set of methods comparing to numpy. These methods have higher executiing efficiency and could be further accelerated by GPU if necessary.
We gave an intuitive introduction to the low-level API in the previous sections, and we will emphasize the introduction on the tensor operation and Autograph.
Tensor operation can be devided into two sub-categories: the structural operation and the mathematical operation.
The structural operation includes tensor creation, indexing and slicing, dimension transformation, combining & splitting, etc.
The mathematical operation includes scalar operation, vector operation, and matrix operation. We will also introduce the broadcasting mechanism of tensor operation.
For the part of Autograph, we will cover its suggested rules, its mechanisms Autograph
and tf.Module
.
Please leave comments in the WeChat official account "Python与算法之美" (Elegant Python and Algorithms) if you want to communicate with the author about the content. The author will try best to reply given the limited time available.
You are also welcomed to reply 加群(join group) in the WeChat official account to join the group chat with the other readers.
此处可能存在不合适展示的内容,页面不予展示。您可通过相关编辑功能自查并修改。
如您确认内容无涉及 不当用语 / 纯广告导流 / 暴力 / 低俗色情 / 侵权 / 盗版 / 虚假 / 无价值内容或违法国家有关法律法规的内容,可点击提交进行申诉,我们将尽快为您处理。