Show HN：我们将31B视频+音频扩散模型压缩为14B，6步，INT8

Show HN: We compressed a 31B video+audio diffusion model to 14B, 6 steps, INT8