Original address machine learning in the Cloud, with TensorFlowWednesday, MarchPosted by Slaven Bilac, software Engineer, Google analytics
machine learning in the cloud with TensorFlow
at Google, researchers collaborate closely and product teams, applying the latest advances in machine learning to Exi Sting products and Services-such asSpeech recognition in the Google app,Search in Google Photos and theSmart Reply feature in Inbox by Gmail-In order to do them more useful. A growing number of Google products is usingTensorFlow, our open source machine learning system, to tackle ML challenges and we would like to enable others do the SAME.
At Google, researchers and product teams work closely together to incorporate the latest advances in machine learning into existing products and services, such as speech recognition in Google Apps, Google photo lookups, and smart recovery features in Gmail inboxes. This is all to make these products and services more practical. The number of Google products using TensorFlow (our open-source machine learning system) is increasing, and in order to take control of machine learning challenges, we will ensure that more products are used TensorFlow.
Today, atGCP NEXT 2016, weAnnounced the alpha release ofCloud Machine Learning, a framework for building and training custom models to is used in intelligent applications.today, at GCP NEXT 2016, we announce the official release of the alpha version of Cloud machine learning, a framework that will be used to build and train customer models that will be applied to AI applications.
Machine Learning projects can come in many sizes, and as we ' ve seen with our open source offeringTensorFlow, projects often need to scale up. Some small tasks is best handled with a local solution running on one's desktop, while large scale applications require B Oth the scale and dependability of a hosted solution. GoogleCloud Machine Learningaims to support the full range and provide a seamless transition from local to cloud environment.
Machine learning projects can be of various sizes, like the open source TensorFlow we have seen, projects usually need to be scaled up. Some small projects that run on-premises solutions on personal computers are easiest to master, while large-scale applications require larger scale and hosted-dependent solutions. Google's cloud machine learning goal is to support a full-area solution and provide a seamless transition from on-premises to cloud environments.
theCloud Machine Learningoffering allows users to run custom distributed learning algorithms based onTensorFlow. In addition to theDeep learningcapabilities that powerCloud Translate API,Cloud Vision API, andCloud Speech API, we provide easy-to-adopt samples for common the tasks like linear regression/classification with very fast convergence p Roperties (based on theSdcaalgorithm) and building a custom image classification model with few hundred training examples (based on theDeCAFalgorithm).
Cloud machine Learning provides the ability to allow users to run distributed learning algorithms based on TensorFlow. Above the capacity of deep learning, enhanced cloud Translate Api,cloud Vision API, and cloud Speech API, we provide some easy to use forin a common taskExamples, such as: Linear regression/classification with very fast convergence properties (based on SDCA algorithm) and a customer image classification model with hundreds of training examples
(based on the decaf algorithm).
We is excited to bring the best ofGoogle Analytics toGoogle Cloud Platform. Learn more about this release and more from GCP Next to theGoogle Cloud Platform Blog.We are excited to bring the best content of Google research to the Google Cloud platform. Want to know more about this release and GCP Next 2016 more content, can go to the Google Cloud Platform blog.
TensorFlow Blog Translation--machine learning in the cloud with TensorFlow
