OSError: stabilityai/sd-vae-ft-mse does not appear to have a file named config.json.

今天在用DIThttps://github.com/facebookresearch/DiT/tree/main

仓库的使用遇到了这个bug，原因是hugging face在加载线上仓库的时候遇到了问题，

解决办法：

先从网上把对应的权重下载到本地：

下载链接：

https://huggingface.co/stabilityai/sd-vae-ft-mse

然后拉一下hugging face的diffusers的官方git仓库

GitHub - huggingface/diffusers: 🤗 Diffusers: State-of-the-art diffusion models for image and audio generation in PyTorch

 然后运行仓库下面scrpits/convert_vae_pt_to_diffusers.py

python convert_vae_pt_to_diffusers.py --vae_pt_path vae-ft-mse-840000-ema-pruned.ckpt --dump_path 文件夹

输入路径就是你刚才下载的权重，输出路径是json信息的保存路径

然后你把报错位置的代码从

vae = AutoencoderKL.from_pretrained(f"stabilityai/sd-vae-ft-{args.vae}").to(device)

替换成就可以了

vae = AutoencoderKL.from_pretrained(pretrained_model_name_or_path='刚才制指定的文件夹').to(device)