Tag: parameter efficient fine tuning

Why it matters: Cut AI training energy without losing accuracy: compare mixed precision, quantization, pruning, distillation, and carbon aware scheduling with real numbers.