yahoo/TensorFlowOnSpark
Distributed deep learning training framework for running TensorFlow programs on Apache Spark and Hadoop clusters.
Not currently ranked — collecting fresh signals.
star history
TensorFlowOnSpark enables distributed TensorFlow training and inference across GPU and CPU servers in Spark/Hadoop clusters. It provides a Spark-compatible API that manages the TensorFlow cluster lifecycle including startup, data ingestion (with options for native TensorFlow file reading or Spark RDD feed), and graceful shutdown. The project minimizes code changes needed to migrate existing TensorFlow programs to a shared grid environment.
Frequently asked
- What is yahoo/TensorFlowOnSpark?
- Distributed deep learning training framework for running TensorFlow programs on Apache Spark and Hadoop clusters.
- Is TensorFlowOnSpark open source?
- Yes — yahoo/TensorFlowOnSpark is open source, released under the Apache-2.0 license.
- What language is TensorFlowOnSpark written in?
- yahoo/TensorFlowOnSpark is primarily written in Python.
- How popular is TensorFlowOnSpark?
- yahoo/TensorFlowOnSpark has 3.8k stars on GitHub.
- Where can I find TensorFlowOnSpark?
- yahoo/TensorFlowOnSpark is on GitHub at https://github.com/yahoo/TensorFlowOnSpark.