- Python 3.6+
- PyTorch 1.0+
- CUDA 11.0+
- Windows/Linux OS for PyTorch GPU
Create and activate a new pip environment
python3 -m venv project
source pytorch/bin/activate
Install PyTorch for pip environment
pip3 install torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/cu113
Install PyTorch GPU in a new conda environment
conda create -n project pytorch torchvision torchaudio cudatoolkit=11.3 -c pytorch
Densenet
DLA
-> checkpoint_teacher : stores the teacher models checkpoints
-> checkpoint : stores the resnet student checkpoint model
-> teacher_models : stores the teacher model specs
# [OPTIONAL:] Train the teacher:
python train_teacher.py --teacher "densenet"
python train_teacher.py --teacher "dla"
or you can just use the checkpoints provided already :)
# You can train the student directly with:
python train_student.py --teacher "densenet"
python train_student.py --teacher "dla"