The client side for MLStorage. MLStorage is an application for running machine learning experiments and storing results as well as generated files, all accessible with a dashboard.
The server side: MLStorage Server.
pip install git+https://github.com/haowen-xu/mlstorage-client.gitmlrun -s http://<ip>:8080 -- python train.py data.csvThe above command runs train.py in the experiment storage directory
(i.e., working directory) assigned by the server http://<ip>:8080.
The script file, train,py, will be copied to the storage directory.
Meanwhile, the data file, data.csv, will be linked to that directory
(on Unix, a symbolic link is created; on Windows, a copied file is created).
The name of the experiment will automatically chosen to be train.py.