A Lab is basically a project that holds code, describes it and keeps track of the code executions along with their logs and outputs.
Think of a lab as a Pen in CodePen or a Fiddle in jsFiddle with the main difference that the whole experience is designed for Machine Learning code which needs to run in the cloud on optimized hardware and were executions are expected to run over many hours or days and produce valuable assets that we want to preserve and make use of.
To create a fresh lab, click the New button in the upper right corner. You can either choose to create a blank lab or start from a template.
Coding is more fun when it’s shared. To use an existing lab as a starting point you can simply fork it to make a copy your own.
Just click Fork in the upper right corner.
It’s good practice to give your lab a descriptive name and a short description. You can also define tags to make it easier for others to find your lab based on a specific tag such as classification or LSTM.
Name, description and tags can be specified from the lab’s settings dialog which opens by clicking on the gear icon next to the lab name.
If for some reason, we don’t want to keep our lab anymore, we can delete it from the Advanced tab of the settings dialog.
In order to prevent us from accidentially deleting a lab, we have to enter the lab name before the Delete button becomes active.