One thing students can do to manage their time is to only switch to GPU mode when they’re training the network. Lots of the steps (pre-processing, etc.) can be run and debugged in CPU mode with unlimited time before saving and switching on GPU mode to train the model.
Articles in this section
- I'm enrolled in a Nanodegree program, how do I contact my team? (School of AI)
- How long will I have access to Nanodegree program content after my term ends?
- Do you have discounts for students?
- How do I manage my GPU hours?
- How do I get an extension?
- Can I still complete the program if I miss a project deadline?
- Where do I go for help with content and project-related questions?
- How do I report a bug in the classroom?
- Who should I contact about billing questions or concerns?
- How will I know when I am ready to graduate?