H2O Quick Start with Flow
Follow this video to start H2O 3.0 with Flow. H2O 3.0 is for enterprise and open-source use, with easy installations for Spark, Python, R, Yarn, Hadoop 1, Amazon EC2, Maven, laptops and standalone clusters. Users can now utilize H2O Flow, a notebook-style graphical user interface with command-line computing, to access fast, scalable H2O algorithms.
Talking Points:
Speakers:
Amy, Unknown, H2O
Read the Full Transcript
Amy:
The following is a quick start guide for launching a local H2O instance. The only prerequisite for the following tutorials is a 64-bit installation of Java version 1.6 or newer.
Getting Started
to start, navigate to our website at h2o.ai on a web browser. Click the download button, which will land you on the downloads page with a listing of all our latest releases. Scroll down to H2O-Dev and choose the latest H2O-Dev stable release. Hit download H2O. Once your download has finished, open up your terminal or command prompt and CD to your downloads folder. Unzip the H2O installation file that you’ve just downloaded, and then CD into the directory that’s just been created.
Launching the JAR File
To launch the jar file that is sitting in the directory, execute and run Java -jar H2O.jar. By default, it should only use about one fourth of the total memory available on your machine. Once the instance is finished launching, you will get a printout of the cloud size and the IP address’s import of all the nodes.
H2O Web UI: Flow Demo
In this case, the singular instance is located at localhost:54321. So point your browser to localhost:54321. From here you have access to H2O’s web UI called Flow. To get a quick demo in flow, scroll down to the help menu on the right hand side. Hit “Browse installed packs,” “Examples,” and load the deep learning example. This is a canned demo of building a deep learning model on the amne data set. You can run through the entire script by hitting the “Run all” button or going to “Flow” and “Run all.” and you have just finished running your first Flow.