A General Distributed Deep Learning Platform


DeepSpark allows distributed execution of both Caffe and Google's TensorFlow deep learning jobs on Apache Sparkā„¢ cluster of machines. It is designed to make large-scale parallel distributed deep learning jobs easy and intuitive for developers and data scientists.

Find Out More

Why DeepSpark?


Fast

DeepSpark has overcome high communication overhead using asynchrony. It is the fastest open source distributed deep learning platform on commodity clusters.

Scalable

DeepSpark has the seamless integration of scalable data management capability of Apache Spark.

Flexible

DeepSpark supports Caffe and Google's TensorFlow. It is the first attempt to integrate TensorFlow with Apache Spark

Up to Date

DeepSpark is an open source project. It is freely available on Github.


Clone DeepSpark on Github

Supporters


Contacts


Give us a call or send us an email and we will get back to you as soon as possible!

+82-2-880-1401