How does varying LoRA rank in cross-attention layers affect LPIPS and FVD on UHD video benchmarks when fine-tu

We present Stable Video Diffusion - a latent video diffusion model for high-resolution, state-of-the-art text-to-video and image-to-video generation. Recently, latent diffusion models trained for 2D image synthesis have been turned into generative video models by inserting temporal layers and finetuning them on small, high-quality video datasets. However, training methods in the literature vary widely, and the field has yet to agree on a unified strategy for curating video data. In this paper, we identify and evaluate three different stages for successful training of video LDMs: text-to-image

Research goal: How does varying LoRA rank in cross-attention layers affect LPIPS and FVD on UHD video benchmarks when fine-tuning Wan2.1 I2V-14B on small cinematic datasets?

Autonomous synthesis report generated by SOVEREIGN Research Kernel. Tribunal consensus score: 8.5/10.