DiffusionBlocks: Training Neural Networks One Block at a Time(pub.sakana.ai)4 points·by sebg·เดือนที่แล้ว·1 commentspub.sakana.aiDiffusionBlocks: Training Neural Networks One Block at a Timehttps://pub.sakana.ai/diffusionblocks/1 commentsPost comment[–]billconan·เดือนที่แล้วreplyI do not understand.how is this different from building smaller transformer layers, and each layer just denoises less?
how is this different from building smaller transformer layers, and each layer just denoises less?