page_type | languages | products | description | |||
---|---|---|---|---|---|---|
sample |
|
|
Learn how to train models using low-priority AML compute. Low-priority compute costs significantly less than dedicated compute, but jobs on low priority compute can be preempted. |
Learn how to train models using low-priority AML compute. This example uses PyTorch to train a model to recognize handwritten digits.