Here is the code for some of the experiments I did with variational autoencoders on multi-modality and atari video prediction. The code uses Torch7 which can be installed here
- All experiments were run on GPU with the following libraries
- cuda/7.5
- cuDNN/v4
- hdf5
- nccl
- nn. Building neural networks.
- nngraph. Building graph-based neural networks.
- optim. Various gradient descent parameter update methods.
- cunn. Provides CUDA support for
nn
. - cudnn. Provides CUDNN support for
nn
. - torch-hdf5. HDF5 interace for torch.
- lfs. Luafilesystem for file manipulation.
- penlight. Commandline argument parser.
- image. Provides support for reading images.
- threads. For multi-threaded data loading.