Knet.jl is a deep learning package implemented in Julia, so you should be able to run it on any machine that can run Julia. It has been extensively tested on Linux machines with NVIDIA GPUs and CUDA libraries, and it has been reported to work on OSX and Windows. If you would like to try it on your own computer, please follow the instructions on Installation. If you would like to try working with a GPU and do not have access to one, take a look at Using Amazon AWS or Using Microsoft Azure. If you find a bug, please open a GitHub issue. If you don't have access to a GPU machine, but would like to experiment with one, Amazon Web Services is a possible solution. I have prepared a machine image (AMI) with everything you need to run Knet. Here are step-by-step instructions for launching a GPU instance with a Knet image (the screens may have changed slightly since this writing).
Features
- Knet machine images are available for AWS, Singularity and Docker
- Define, train and test the LeNet model for the MNIST handwritten digit recognition
- Knet is an open-source project and we are always open to new contributions
- Deep learning framework implemented in Julia by Deniz Yuret
- It supports GPU operation and automatic differentiation
- Dynamic computational graphs for models
Project Samples
