• 8GB GPU for training. 4GB GPU for evaluation.
• CUDA & Nvidia driver (download & install: https://developer.nvidia.com/cuda-downloads)
• CuDNN (download & install: https://developer.nvidia.com/cudnn)
• Prerequisites (sudo apt-get install -y gcc g++ gfortran build-essential git wget libopenblas-dev python-dev python-pip python-nose python-numpy python-scipy x264 ffmpeg)
• Theano (sudo pip install --upgrade --no-deps git+git://github.com/Theano/Theano.git@5159a6b3cc2875f90e86449e5b3914f7bcb76452)
• Lasagne (sudo pip install --upgrade --no-deps git+git://github.com/Lasagne/Lasagne.git)
• bcolz (sudo pip install bcolz)
• moviepy (sudo pip install moviepy)
• OpenCV (install: https://github.com/jayrambhia/Install-OpenCV)

Set your custom paths in paths.py.

This takes more or less 150GB.

python scripts/videos_prep.py

This is not required as the pretrained models are included.

THEANO_FLAGS=device=gpu python train.py --jobs=4 --config=models/3drnn_sep2d_elu_nod_mlr.py

THEANO_FLAGS=device=gpu python predict.py --set=test --meta=metadata/3drnn_sep2d_elu_nod_mlr-hond-20170620-115014.pkl

python submission.py --set=test --meta=metadata/3drnn_sep2d_elu_nod_mlr-hond-20170620-115014.pkl

Replace "test" to "valid" to make a submission file for the validation set. If the models are retrained, replace the metadata files to the new ones.