Closed alimanfoo closed 3 months ago
It is possible to figure it out from the references but I think it might help to have a list of PRs addressing this in order:
W1M1: probably #195 (merged)
W1M2: probably #195 (merged)
W1M3: probably #195 (merged)
W1M4: probably #195 (merged)
W2M1: probably #201 (merged)
W2M2: No package is installed
W2M3: probably #201 (merged)
W2M4: probably #201 (merged)
W3M1: No package is installed
W3M2: No package is installed
W3M3: probably #210 (merged)
W3M4: probably #210 (merged)
W4M1: No package is installed
W4M2: No package is installed
W4M3: probably #218 (merged)
W4M4: probably #218 (merged)
W5M1: #239 (merged)
W5M2: No package is installed
W5M3: probably #228 (merged)
W5M4: #244 (merged)
W6M1: #229 (merged)
W6M2: No package is installed
W6M3: #231 (merged)
W6M4: #230 (merged)
W7M1: #232 (merged)
W7M2: #240 (merged)
W7M3: #243 (merged)
W7M4: #260 (merged)
W8M1: No package is installed
W8M2: No package is installed
W8M3: No package is installed
W8M4: No package is installed
All branches have been merged.
Jupyter notebooks generally support a
%pip
line magic command, which runs pip and is guaranteed to install packages into whatever environment is being used to run the current kernel.The more general
!pip
can be used to run pip as a system command, which will install packages into the base environment.On colab, either works the same. However, on Amazon Sagemaker Studio Lab,
%pip
is required to ensure packages are installed in the current environment. This is probably a generally safer option.For future compatibility with sagemaker lab, suggest we switch over notebooks here to using
%pip
.