The presentation were created in the scope of a project on tensorflow and Google Cloud Platform together with Fabien Tarrade, please visit github/tarrade/proj_DL_models_and_pipelines_with_GCP for latest developments.
Slides for
- Windows Users, underlying notebook (in nbviewer)
- Linux/Mac Users, view underlying notebook (in nbviewer)
Follow me on 
.
Install Anaconda Distribution on your system. This provides you with the most important packages on your system.
Install Jupter Notebook extension into base environment.
conda activate base
conda install -c conda-forge jupyter_contrib_nbextensions #triggers system-wide installation- go to the directory in which you want to download the package from git
- download the package from Github:
git clone https://github.com/enryH/gcp_ml_engine_talk/- or download zipped repository.
- open a command line in the directory that contains the code from GitHub.
This will provide you a virtual environment of python packages needed to run the code.
-
create a python env based on a list of packages from environment.yml
conda env create -f environment.yml -n gcp_dl -
activate the env
conda activate gcp_dl
In case of issue clean all the cache in conda
conda clean -a
-
In case you want to use nbdime for checking differences in notebooks:
nbdime config-git --enable --global -
In case you want to use interactive slides using RISE:
conda activate gcp_dl conda install -c conda_forge rise
-
update a python env based on a list of packages from environment.yml
conda env update -f environment.yml -n gcp_dl -
delete the env to recreate it when too many changes are done
conda env remove -n gcp_dl
- open a command line
- go to the directory in which you want to download the package from git
- activate the env:
conda activate gcp_dl - execute TensorBoard:
tensorboard --logdir ./results - open a web browser (Firefox) and copy the link that appear in the prompt above:
http://<something>:6006
