We’re excited to announce the newest version of our Bitfusion Ubuntu 14 Theano AMI is now available in the AWS Marketplace.
Theano is python library that provides that provides tight integration with NumPy for numerical computing and has a large amount of community use and support. Theano is great for machine learning and deep learning use cases since it can leverage CPUs or GPUs transparently, and it can quickly build up optimized symbolic computational graphs with gradients automatically computed. Theano is also a mature framework that has been powering large-scale computationally intensive scientific investigations since 2007.
Recently, AWS announced their next generation GPU P2 instances, which provide up to 16 NVIDIA K80 GPUs, 64 vCPUs and 732 GiB of host memory. This is a welcome relief, given that the previous generation GPU G2 instances were around for a while and starting to show their age. Our latest Theano AMI has updated NVIDIA drivers, the CUDA toolkit added, and CUDNN support, allowing you to tap into these new powerful instances.
Regardless of whether you prefer Python 2 or Python 3, we’ve added new and updated tools, libraries, and frameworks to support and handle both.
When working with Theano, there are many libraries that provide complementary functionality, or layer on top of Theano for additional capabilities. In order to provide a truly comprehensive data science and deep learning toolset, we’re including them all by default.
Lasagne is a lightweight Python library for building and training neural nets on Theano. It supports Convolutional Neural Networks (CNNs), recurrent networks including Long Short-Term Memory (LSTM), or combinations of the two, multiple inputs and multiple outputs including auxiliary classifiers, a variety of optimization methods, and other useful features for rapid, productive work with neural nets.
Keras is an expressive, high-level deep learning library for Theano (and TensorFlow). Keras also allows building and training neural nets, but focuses on experimentation and rapid prototyping for faster development iterations. Normally when you install Keras, it defaults to using TensorFlow as the backend, but we’ve made sure you don’t have to worry about reconfiguring things -- the Theano AMI has Keras talking to Theano out of the box.
Pandas is library of data structures and data analysis tools that extend Python’s data munging and prep strengths with powerful data analysis and modeling capabilities.
Scikit-Learn is a machine learning library built on NumPy, SciPy and matplotlib for data mining and data analysis that supports a variety of different models and techniques.
PyCUDA is rapidly becoming the leading way to interact with NVIDIA CUDA from Python due to its effective handling of memory management and overall convenience and comprehensiveness. By bundling this capability with the ability to compile CUDA Kernels directly from Jupyter notebooks, you can have the ability to interact with the CUDA parallel computation API at a very deep level while also simplicity and ease of use of a notebook user interface.
Try out these new additions and enhancements in the AWS marketplace, all our AMIs come with a 5 day free trial: Bitfusion Ubuntu 14 Theano AMI.