From the course: Apache Spark Deep Learning Essential Training

Unlock the full course today

Join today to access over 22,400 courses taught by industry experts or purchase this course individually.

Deploying models as SQL functions

Deploying models as SQL functions - Spark DataFrames Tutorial

From the course: Apache Spark Deep Learning Essential Training

Start my 1-month free trial

Deploying models as SQL functions

- [Instructor] Now one way to take a deep learning model to production, is to deploy it as a Spark SQL UDF or User-defined Function. This means that anyone who knows SQL can use it. How cool is that? So let's head over to our Notebook. I select Workspace, I select the Down Arrow, I select Import, File, Browse, Exercises, 05_01... _begin, and I select Import. Now we need to check we have the correct versions of the libraries installed, including the Deep Learning Library. So attach the notebook to the cluster. We're going to use the flower dataset again, but if you've downloaded it to your cluster you don't have to do that again. So you've got the directories daisy, dandelion, roses, sunflowers, and tulips under the flower_photos directory. And this time we're going to create a directory called sample_images and copy three of the daisy images across, and we're going to load the three images of the daisy into the dataframe. Now in the dataframe you can see the column named Image, with…

Contents