Developing for Hub is as you expect for most Python packages.
It is highly recommended you use an environment manager when developing for any project, Hub is no exception. For more information on how conda works, please visit their website. You can create a new conda environment with the following (replace [version] with your version, must be greater than python3.6)
conda create -n hub python=[version]
Activate the environment
conda activate hub
Note: If you don’t like using conda, you can use venv instead.
When installing the package, make sure you use the
-e flag, this allows you to
import hub from another project (with the same conda/venv environment) & make changes directly to the source with immediate updates.
git clone https://github.com/activeloopai/Hub cd Hub pip install -e .
Make sure that, after following the preceeding steps, you can now run this python script
from hub import Dataset mnist = Dataset("activeloop/mnist") # loading the MNIST data lazily # saving time with *compute* to retrieve just the necessary data mnist["image"][0:1000].compute()
If you receive the error
botocore.exceptions.ClientError: An error occurred (InvalidAccessKeyId) when calling the GetObject operation: The AWS Access Key Id you provided does not exist in our records.
Double check that you included
aws_session_token within your
~/.aws/credentials profile. More information regarding AWS credentials can be found here.
If you are having a related problem, please read/comment on the reference issue or you can setup with docker.