nainstalovanie python packagov je potrebne mat pip a python aspon 3.5
pip by mal byt normalne nainstalovany ak nie je, podla verzie linuxu nainstaluj https://packaging.python.org/guides/installing-using-linux-tools/
pak pre kazdy balicek napises sudo pip3 install jmeno balicku
dynet: http://dynet.readthedocs.io/en/latest/python.html pre cpu verzi pip3 install dynet pre gpu a najnovejsi verzi BACKEND=cuda pip install git+https://github.com/clab/dynet#egg=dynet
networkx: to je package na spracovanie grafov https://networkx.github.io/documentation/stable/install.html pip install neetworkx
pytorch: pytorch ma problem s poliami/listami premenych ma sa na to pouzivat specialny modul ale ten nie je rychly spis pomaly, co je velky problem ak chcem nacitat nahodne grafy s roznym poctom vrstiev, ale ak sa to napise staticky tak je to fakt rychle https://pytorch.org/ uz v zaklade ma gpu pip3 install torch torchvision
tensorflow: https://www.tensorflow.org/install/install_linux sudo apt-get install cuda-command-line-tools pak pridas vygenerovane cuda kniznice do cesty pre kniznice xport LD_LIBRARY_PATH=${LD_LIBRARY_PATH:+${LD_LIBRARY_PATH}:}/usr/local/cuda/extras/CUPTI/lib64 pak pre gpu pip3 install tensorflow-gpu pak pre cpu pip3 install tensorflow
pre pouzivanie c++ potrebujes eigen na to aby to fungoval musis stiahnut a extractovat kody pomocou hg clone https://bitbucket.org/eigen/eigen/ nebo git clone https://github.com/eigenteam/eigen-git-mirror a nalinkovat do /usr/include, pak by to malo fungovat