Accelerate video diffusion inference with cache, sparse attention, token pruning, quantization, and kernel fusion.
Included