koike-ya / ml_dl_pkg

machine learning packages

What's this repository for?

To boost baseline model construction and incremental hypothesis testing speed.

You can use this library for both machine learning and deep learning tasks only with changing some argments.

Requirements

• cuda >= 10.0

Setup Environment

Docker

To use the image with a GPU you'll need to have nvidia-docker installed.

sudo docker run -ti --gpus all -v `pwd`/data:/workspace/data -p 8888:8888 --net=host --ipc=host pelada/ml_pkg:latest # Enters with /bin/bash, mounting the /data drive in the container

From source

conda create -n ml_dl_pkg python=3.7
source activate ml_dl_pkg

Only for mac user, Please install xgboost from source

https://xgboost.readthedocs.io/en/latest/build.html#building-on-osx

Apex(for GPU users)

ref: https://github.com/NVIDIA/apex

cd ../
git clone https://github.com/NVIDIA/apex;cd apex
pip install -v --no-cache-dir --global-option="--cpp_ext" --global-option="--cuda_ext" ./
cd ../ml_dl_pkg

For CPU users,

mkdir -p apex/amp 

Lastly,

pip install -r requirements.txt
python setup.py install

Example

mkdir input/eeg;unzip 'input/*.zip' -d input/eeg/;
python example.py --transform spectrogram