TheLastBen / fast-stable-diffusion

fast-stable-diffusion + DreamBooth
MIT License
7.57k stars 1.31k forks source link

The future belongs to a different loop? #2540

Open mmehrle opened 1 year ago

mmehrle commented 1 year ago

Just downloaded the new notebook, so I'm up to date. This error seems to be happening with 1.5 as well SDXL:

Startup time: 204.0s (launcher: 0.8s, import torch: 17.5s, import gradio: 2.0s, setup paths: 22.2s, import ldm: 0.6s, initialize shared: 6.2s, other imports: 32.8s, setup codeformer: 8.8s, setup gfpgan: 0.4s, list SD models: 7.2s, load scripts: 30.8s, load upscalers: 0.8s, reload hypernetworks: 0.6s, initialize extra networks: 20.4s, create ui: 15.6s, gradio launch: 34.1s, add APIs: 3.1s). Creating model from config: /content/gdrive/MyDrive/sd/stable-diffusion-webui/configs/v1-inference.yaml The future belongs to a different loop than the one specified as the loop argument The future belongs to a different loop than the one specified as the loop argument The future belongs to a different loop than the one specified as the loop argument The future belongs to a different loop than the one specified as the loop argument The future belongs to a different loop than the one specified as the loop argument The future belongs to a different loop than the one specified as the loop argument The future belongs to a different loop than the one specified as the loop argument

TheLastBen commented 1 year ago

use a fresh sd folder

mmehrle commented 1 year ago

Meaning, go to my drive and wipe out the current sd folder? That would also delete my existing configs and downloads in stable-diffusion-webui. I would have to replace the default one manually afterward I guess...

TheLastBen commented 1 year ago

don't delete it, just rename it

MangleMike commented 1 year ago

This works for me in Google Colab, it gets rid of the error, but I do have to move all my models and LoRA's etc. The reason is when I go to install my Automatic1111 from my copy of Colab note book it installs a whole new SD folder onto my Gdrive that it is now linked to. When you say just rename it, do you mean we can now just relink the Colab note book to that renamed file in the Gdrive? If so I can't work out how to do it.

iaclaudioia8 commented 1 year ago

When I got this issue, I usually stopped and started again the tab "start Stable Diffusion".

johnaweber commented 11 months ago

I have been able to avoid this error by deleting the base model before starting the notebook again. For example, if the last time I ran the notebook I'd selected the v1.5 model, I would delete the following file before starting:

sd/stable-diffusion-webui/models/Stable-diffusion/v1-5-pruned-emaonly.safetensors

This has allowed me to keep everything else in place.

acronoo commented 9 months ago

Delete just "libtcmalloc" and "stablediffusion" folders solve this problem to me