Which statement describes differences between training and inference in AI learning models?

A: The inference process requires high throughput to enable rapid training on large datasets. B: The training process is more computationally demanding because it processes new data. C: The training process is more optimized for low latency, which enables faster analysis of large datasets. D: The inference process requires continuous model adjustments to maintain accuracy with new data.