Google Cloud AI provides modern machine learning services, with pre-trained models and a service to generate tailored models.
N/A
TensorFlow
Score 8.9 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.
Google's documentation for their AI and Machine Learning products is a bit more straightforward and still much easier to onboard into compared to the Azure Machine Learning and other AI products. Additionally, Google's Cloud AI products provide more comprehensive specific …
Google Cloud AI is a wonderful product for companies that are looking to offset AI and ML processing power to cloud APIs, and specific Machine Learning use cases to APIs as well. For companies that are looking for very specific, customized ML capabilities that require lots of fine-tuning, it may be better to do this sort of processing through open-source libraries locally, to offset the costs that your company might incur through this API usage.
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).
Some of the build in/supported AI modules that can be deployed, for example Tensorflow, do not have up-to-date documentation so what is actually implemented in the latest rev is not what is mentioned in the documentation, resulting in a lot of debugging time.
Customization of existing modules and libraries is harder and it does need time and experience to learn.
Google Cloud AI can do a better job in providing better support for Python and other coding languages.
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.
We are extremely satisfied with the impact that this tool has made on our organization since we have practically moved from crawling to walking in the process of generating information for our main task to investigate in the field through interviews. With the audio to text translation tool there is a difference from heaven to earth in the time of feeding our internal data.
So far we have not had any problems with the application, it is extremely easy to use and it is also easy to install. It does not require much training since the instruction is quite clear for its execution and application in the procedures of the company. The team in charge of this application is very pleased to work with this app.
Every rep has been nice and helpful whenever I call for help. One of the systems froze and wouldn't start back up and with the help of our assigned rep we got everything back up in a timely manner. This helped us not lose customers and money.
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.
In fact, you only need the basic tech knowledge to do a Google search. You need to know if your organization requires it or not,. our organization required it. And that is why we acquired it and solved a need that we had been suffering from. This is part of the modernization of an organization and part of its growth as a company.
These are basic tools although useful, you can't simply ignore them or say they are not good. These tools also have their own values. But, Yes, Google is an advanced one, A king in the field of offering a wide range of tools, quality, speed, easy to use, automation, prebuild, and cost-effective make them a leader and differentiate them from others.
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
Artificial intelligence and automation seems 'free' and draws the organization in, without seeming to spend a lot of funds. A positive impact, but who is actually tracking the cost?
We want our employees to use it, but many resist technology or are scared of it, so we need a way to make them feel more comfortable with the AI.
The ROI seems positive since we are full in with Google, and the tools come along with the functionality.
