Deep Learning Virtual Machine
I assume you already have VirtualBox (version 5+) installed, if you don't, please download and install it.
- Download and install Vagrant if you haven't previously done so.
- Change into the
deep-learning-vmdirectory and run
vagrant up- this creates the VM.
- Deep Learning,
- Keras, 2.0.6
- TensorFlow, 1.3.0
- Theano, 0.8.2
- Python, 3.5.2
- H5py, 2.6.0
- MXNet, 0.10.0
- numpy, 1.11.1
- pandas, 0.18.1
- scipy, 0.18.1
In addition, a Jupyter notebook server is also installed. You can view it from the host's browser at http://localhost:8100. Password : password
This VM comes with a few notebooks (with required datasets).
If you'd prefer the obligatory handwritten digits example,
$ vagrant ssh $ curl -sSL https://raw.githubusercontent.com/fchollet/keras/master/examples/mnist_mlp.py | python
The script above,
mnist_mlp.py, trains a simple deep neural network on the MNIST dataset.
If all goes well, you should see a test accuracy close to 98.40% after 20 epochs.
Testing Theano (Optional)
Create a new notebook, and key in the following
import theano theano.test()
then Run. It took ~2 hours for the 2995 tests to run in the VM on my laptop.