Databricks supports deep learning libraries for building and applying Neural Networks. This section gives examples to get started with deep learning in Databricks.
Graphics Processing Units (GPUs) can accelerate deep learning and other Machine Learning tasks. For more information about creating GPU-enabled clusters, see GPU-enabled clusters.
Init Scripts: To use deep learning libraries, we recommend that users use Init Scripts to install the needed libraries on clusters upon creation.
We provide Init Scripts for several popular libraries, as well as accompanying example notebooks to get started.
TensorFlow and MXNet do not require Init Scripts for their latest releases! Most other libraries, such as Theano, Caffe, etc., should use Init Scripts.
Dependencies: Databricks includes pre-installed GPU hardware drivers and NVIDIA libraries such as CUDA. See GPU-enabled clusters for more information.
- Integrating Deep Learning Libraries with Apache Spark