- This event has passed.
TensorFlow & TensorFrames with Apache Spark + A Deep-dive into Structured Streaming
January 18, 2017 @ 7:00 pm - 9:00 pm
We are proud to announce our first meetup on the 18th of January in collaboration with Databricks. We will have two talks:
Introduction to TensorFlow with Apache Spark using TensorFrames.
TensorFrames provide the bridge between Spark and the famous TensorFlow framework, recently open sourced by Google. Unlike competing alternatives like SparkNet, TensorFrames relies on the DataSet/DataFrame API of Spark and has in-depth knowledge of the memory-efficient representation of the data in Spark, therefore minimizing the memory redundancy between the two frameworks.
After a brief introduction to TensorFlow, the talk will focus on how to enhance the computational power of TensorFlow by leveraging the distributed nature of Spark with TensorFrames.
Speaker bio: Marco Saviano is a Big data Engineer for Agile Lab. He has acquired in-depth knowledge of deep learning techniques during his academic years and has been using TensorFlow since its first release.
A Deep-dive into Structured Streaming
In Apache Spark 2.0, we have extended DataFrames and Datasets in Spark to handle streaming data. Streaming Datasets not only provides a single programming abstraction for batch and streaming data, it brings support for event-time based processing, out-of-order/delayed data, sessionization and tight integration with non-streaming data sources and sinks. In this talk, Burak will take a deep dive into the concepts and the API and show how this simplifies building complex “continuous applications.”
Speaker bio: Burak Yavuz is a Software Engineer at Databricks. He’s been contributing to Spark since Spark 1.1, and is the maintainer of Spark Packages ( https://spark-packages.org,https://spark-packages.appspot.com/).
After the talks beer and snacks will be served 🙂
We hope you will attend and we look forward to meet each and everyone of you.
Merry Christmas and happy new year!!
Marco, Vito & Paolo