Hi Everyone,
We just published a new and quick TensorFlow basics overview page.
Let us know what you think.
5 Likes
I was thinking that some section like “Frequently Asked Questions” could be useful. So that people would search it before asking the same basic question, which was already answered.
It helps to know why a technology was created.An introduction should not leave me reading for pages and pages saying, “but what was the point of all this?”.
A Tensor is not what is interesting about Tensors. Tensors exist to create a chain of Tensors, which represents a series of computations that can be run over and over again. It would help to include a demonstration and maybe an illustration of a chain of two Tensors. Without this concept of a data structure representing a chain, there is no reason for Tensors to exist.
I’m not familiar with this terminology. What do you mean by “chain” here?
“Chain” was the wrong word, apologies.
Tensors are about creating directed graphs of computations. The directed graph represents a series of computations that can be run over and over again. The Tensor data type is an output node of this graph. A computation graph can have more than one output node.
To explain this well requires a diagram as well as words.
Thanks, I thought that might be what you meant, but I just wanted to be sure.