Audio driven推理出来的视频，口型明显不对

[ahqzy]

命令：python -m scripts.audio2vid --config ./configs/prompts/animation_audio2.yaml -W 512 -H 512 -acc

视频：
https://github.com/user-attachments/assets/feb9e3f1-42e9-4e97-9ee8-ce7d4426b33b

[ahqzy]

animation_audio2.yaml 内容：
pretrained_base_model_path: './pretrained_model/stable-diffusion-v1-5'
pretrained_vae_path: './pretrained_model/sd-vae-ft-mse'
image_encoder_path: './pretrained_model/image_encoder'

denoising_unet_path: "./pretrained_model/denoising_unet.pth"
reference_unet_path: "./pretrained_model/reference_unet.pth"
pose_guider_path: "./pretrained_model/pose_guider.pth"
motion_module_path: "./pretrained_model/motion_module.pth"

audio_inference_config: "./configs/inference/inference_audio.yaml"
inference_config: "./configs/inference/inference_v2.yaml"
weight_dtype: 'fp16'

test_cases:
"./configs/inference/ref_images/lyl.png":
- "/home/zqiu/video/gdg_16k.wav"

[ahqzy]

音频文件：
gdg_16k.zip

[fcfchd]

这牙真可怕