TMElyralab / MuseV

MuseV: Infinite-length and High Fidelity Virtual Human Video Generation with Visual Conditioned Parallel Denoising
Other
2.1k stars 214 forks source link

请问。执行 python app.py时 为什么每次都要重新下载模型 #132

Closed SevenMpp closed 1 month ago

SevenMpp commented 1 month ago

您好!请问, 已经下载了git clone https://huggingface.co/TMElyralab/MuseV ./checkpoints, 看着是下载完全.但是为什么执行 python app,py都需要重新下载 diffusion_pytorch_model.safetensors 约为1.45G. 而checkpoints/vae/sd-vae-ft-mse ,stabilityai/sd-vae-ft-mse/vae 和 runwayml/stable-diffusion-v1-5/vae 中的大小也是335M. 并且每次都需要重新在下载一些模型文件,且,每次运行都是重新下载, 请问有此模型的下载地址吗?并且其是替换checkpoints/vae/sd-vae-ft-mse 中的 diffusion_pytorch_model.safetensors吗???

企业微信截图_1717489024131
SevenMpp commented 1 month ago

已解决 关闭