Guided Diffusion

LoRA (Low-Rank Adaptation)

Efficient fine-tuning technique that injects small low-rank matrices into the model's attention layers, allowing learning of new styles or concepts with minimal parameters.

← Indietro