Google’s TensorFlow: Two Heads are better than one
Google has recently announced TensorFlow, which is an open source software library for machine learning, will be available under the Apache 2.0 license.
TensorFlow, according to tensorflow.org, is: “an open source software library for numerical computation using data flow graphs. Nodes in the graph represent mathematical operations, while the graph edges represent the multidimensional data arrays (tensors) communicated between them. The flexible architecture allows you to deploy computation to one or more CPUs or GPUs in a desktop, server, or mobile device with a single API.”
A data flow graph, according to tensorflow.org, describes “…mathematical computation with a directed graph of nodes & edges. Nodes typically implement mathematical operations, but can also represent endpoints to feed in data, push out results, or read/write persistent variables. Edges describe the input/output relationships between nodes. These data edges carry dynamically-sized multidimensional data arrays, or tensors… Nodes are assigned to computational devices and execute asynchronously and in parallel once all the tensors on their incoming edges becomes available.” “The flow of tensors through the graph is where TensorFlow gets its name.”
Open sourcing TensorFlow means that anyone will be able access, and change the software. (hopefully for the better) This may seem like a strange decision, most businesses haven’t become successful by giving away their best ideas, but in this case it makes sense.
Google will effectively be able to speed up the development process, since many people will be able to access it and make changes and improvements simultaneously.
Not only will this allow Google to speed up the progression of TensorFlow, but it will allow them to do it at a lower cost, since it will be happening outside the company.
Hopefully Google’s progressive approach will help foster a community focused on advancement and improvement in the tech industry.