Google officially launches TensorBoard.dev and TensorFlow Enterprise

After much speculation, Google finally announced a preview of TensorBoard.dev, a tool to share the ability to visualize machine learning TensorBoard and TensorFlow Enterprise, a cloud service developed through collaboration. between Google's TensorFlow and Cloud team.

'Now you can store and track your ML (machine learning) tests and share them publicly, without any setup. Simply upload your personal diary and share the URL so others can see the tests and what you're doing with TensorBoard ', Google's VP of engineering, Megan Kacholia.

TensorFlow Enterprise was created with the purpose of bringing an optimized version of TensorFlow open source machine learning framework to customers of large enterprises. The tool is compatible with Google's AI Platform (AI Platform) and Kubernetes Engine, as well as optimized versions of Deep Learning VM and Deep Learning Container. The service is designed to provide 3x more efficient data reading - the result of changes to the way TensorFlow reads and stores files - and supports up to 3 years of patches. security and choose to fix errors.

In addition, TensorFlow Enterprise is also likely to come with a direct support package from Google Cloud and TensorFlow engineers to help businesses to effectively train and deploy their own AI systems.

In related news, last week, the Google Cloud Platform introduced updates to AI Platform, a cloud-based software that supports data collaboration between data scientists to create new models. AI figure. Last month, Google launched TensorFlow 2.0 with tighter integration of Keras.