Yesterday we looked at Training Job structure. Today we see the built-in algorithms you can run inside them. Built-in algorithms are pre-built containers from AWS—you don't write a single line of training code; just provide data in the right format and training happens. It's a choice between writing PyTorch from scratch or leveraging AWS-optimized algorithms.
In the MLA-C01 exam, built-in algorithms appear as keywords like "which algorithm for which problem", "input format (RecordIO vs CSV)", "characteristics of each algorithm". You don't need to memorize every algorithm; the key is problem-type mapping with representative algorithms and format rules.