Users can build custom conversational experiences using Google Assistant’s voice and visual APIs. Take users on journeys through a product, using Assistant’s natural language understanding (NLU) capabilities and developer tools.
N/A
TensorFlow
Score 7.7 out of 10
N/A
TensorFlow is an open-source machine learning software library for numerical computation using data flow graphs. It was originally developed by Google.
I'm in a Me vs. The World environment rather often. I can connect to my outer realm when heading to live meetings. Auditions, job assignments all via my assistant. I like having the ability to capture the moment and rewrite it as well. This is a primary driver for me. Sometimes branching out or when collaborating, I think I work a little harder in the moment than Google Assistant might but that is moreso my limitations and not the feature so much. I catch this scene when I'm in a group environment or at times having to create and respond to a larger scale event. Not a deal breaker for me however.
TensorFlow is great for most deep learning purposes. This is especially true in two domains: 1. Computer vision: image classification, object detection and image generation via generative adversarial networks 2. Natural language processing: text classification and generation. The good community support often means that a lot of off-the-shelf models can be used to prove a concept or test an idea quickly. That, and Google's promotion of Colab means that ideas can be shared quite freely. Training, visualizing and debugging models is very easy in TensorFlow, compared to other platforms (especially the good old Caffe days). In terms of productionizing, it's a bit of a mixed bag. In our case, most of our feature building is performed via Apache Spark. This means having to convert Parquet (columnar optimized) files to a TensorFlow friendly format i.e., protobufs. The lack of good JVM bindings mean that our projects end up being a mix of Python and Scala. This makes it hard to reuse some of the tooling and support we wrote in Scala. This is where MXNet shines better (though its Scala API could do with more work).
I think newer, complementary ideas are a bit sharper than Google Assistant especially in a Q&A environment or when seeking some depth to a subject. That enhancement is to be expected I feel. And Google Assistant is not so self limiting so I don't have a lot of improvement needs because I use this for what I've become accustomed to and for the ability overall.
It is always important to do your best around hectic places, in bad tower signal areas or even if trying to do something new while using Google Assistant. Have patience in the setting. It pays off.
Theano is perhaps a bit faster and eats up less memory than TensorFlow on a given GPU, perhaps due to element-wise ops. Tensorflow wins for multi-GPU and “compilation” time.
I feel this can be adjusted and after some trial and error you sort of start knowing what will work and how. And I have to say the overall impact becomes personal and we are all different. I'm small scale and as I've said, it works.
Community support for TensorFlow is great. There's a huge community that truly loves the platform and there are many examples of development in TensorFlow. Often, when a new good technique is published, there will be a TensorFlow implementation not long after. This makes it quick to ally the latest techniques from academia straight to production-grade systems. Tooling around TensorFlow is also good. TensorBoard has been such a useful tool, I can't imagine how hard it would be to debug a deep neural network gone wrong without TensorBoard.
I chose this because it was easier for me and can be accessed via mobile and laptop too because it enables cross device support because it helps in adding more depth to my life, and can help me save tons of time.
Keras is built on top of TensorFlow, but it is much simpler to use and more Python style friendly, so if you don't want to focus on too many details or control and not focus on some advanced features, Keras is one of the best options, but as far as if you want to dig into more, for sure TensorFlow is the right choice
