AWS instructions
Starting up AWS
- Go to our MLSpace webpage, and sign on with your MLSpace-specific password (your username is m<alpha>).
- Under “My Resources”, click “Projects”, and choose the appropriate project.
- On the menu in the left margin, under the project name, choose “Notebook instances”.
- If making a new notebook, click “Create notebook instance”
- Give it a name which includes your alpha.
- For “Notebook instance type,” use
ml.g4dn.xlargeunder “Accelerated Computing”. - Click “Create notebook instance.”
- Wait a couple minutes. You may be directed to a page saying “not found” - that’s OK, just click “Notebook instances” on the left margin.
- If restarting a previous notebook, click the radio button, then under “Actions” choose “Start”. Wait for the status to become green.
- Once the Status is green under Notebook instances, you are spending money.
- Click the radio button next to the instance name, and choose
Open JupyterLab. - From the Jupyter Lab, you can upload an .ipynb file by clicking the up arrow icon that appears approximately below
Run.
Closing down AWS
Once you’re done, you must turn your notebook instance off, or it continues to cost money. Within the “Notebook instances,” click the radio button next to the notebook, and click Actions -> Stop.