Error when attempting to install dependencies from conda environment.yml about ml-workspace HOT 1 CLOSED

@tbrassel The Install Dependencies section is related to running code via the workspace container without starting any interactive tools. In the normal usage, the environment.yml is currently not installed automatically (we might consider that in the future). When you start the workspace, it is somewhat like a virtual machine running on you computer with its own OS, installed libraries and tools.

The beste way to install additional dependencies is by just using the terminal within the workspace (find out more in the extensibility section). So, you can just create a new conda environment from your environment.yml as explained here. VS-code should automatically discover this new environment. If you like to use Jupyter, just use ipykernel to install a kernel in the given environemnt as explained here.

However, since the base conda environment of the workspace already comes preinstalled with many of the most commonly used libraries, you might not even have to install your environment.yml (depending on the complexity of your dependencies). I mostly manage my project dependencies within a requirements.txt which I install via pip within the base conda environment of the workspace.

from ml-workspace.