Deep Learning environment setup
Aug 9, 2017 · 1 min read
Why?
Last March I started the Deep Learning Nanodegree program Udacity. I took it after I had some practice w Udacity’s Free Machine Learning course and even implemented some classifiers on @Kaggle. I played with regressions, SVM, Decision Trees etc. All this time I knew that eventually I will want to use DL because of its superior performance and amazing use cases I read about. This is how I created my setup:
Start Here
- Download Miniconda
2. Install form terminal:
bash Miniconda2-latest-MacOSX-x86_64.sh3. add conda path to your .bashrc/.zshrc:
export PATH=”~/miniconda2/bin:$PATH”4. Now, in order to start working with python in your environment, run in terminal:
conda create -n dlnd python=3source activate dlnd #activate conda envconda install numpy matplotlib pandas jupyter notebook bokeh h5py tqdm tensorflow ipykernel -y # install packages for this env aloneipython kernel install --name dlnd #link the ipython interpreter to this envjupyter notebook — ip=’*’ — port=8888 — no-browser — allow-root
and you’re good to go. open in the browser the link jupyter provided.
cheers!
