DiTFastAttn: Attention Compression for Diffusion Transformer Models Paper • 2406.08552 • Published Jun 12 • 23