search

ShivamShrirao / diffusers

🤗 Diffusers: State-of-the-art diffusion models for image and audio generation in PyTorch
https://huggingface.co/docs/diffusers
Apache License 2.0
1.89k stars 507 forks source link

recently can't use outputted ckpt in automatic1111 #150

Open habanerogossip opened 1 year ago

habanerogossip commented 1 year ago

Describe the bug

I've been happily using this repo to train dreambooth models and then convert them to ckpt to use in automatic1111's UI. It worked fine a couple days ago, but today when I put the ckpt into automatic1111 I get an error that the file is bad and might contain malicious code. On windows 10 it shows that all my past ckpt files had a specific file size of 2,082,868KB and now all my new ones have a slightly different file size of 2,082,805KB, so they definitely differ in some way. What could be the cause of this issue?

Reproduction

No response

Logs

No response

System Info

running dreambooth on A5000 and automatic1111 UI on windows 10

maxlund commented 1 year ago

If you want to load it anyway, there is a flag you can use as argument in automatic1111 webui launch.py: --disable-safe-unpickle Can be set in webui-user.sh (or .bat equivalent) e.g. COMMANDLINE_ARGS="--xformers --opt-split-attention --disable-safe-unpickle"