Help with error

[aeon3]

Hey im trying to install it but I ran into this issue:

huggingface_hub.utils._errors.RepositoryNotFoundError: 401 Client Error: Repository Not Found for url: https://huggingface.co/api/models/CompVis/stable-diffusion-v1-4/revision/fp16. If the repo is private, make sure you are authenticated. (R

I have no clue what do to. I already created a new token with Read permissions and I applied it like this:

docker run --name stable-diffusion --gpus all -it -e TOKEN= -p 7860:7860 nicklucche/stable-diffusion

Any help for coding illiterate?

[NickLucche]

Hi, did you register to this repo https://huggingface.co/CompVis/stable-diffusion-v1-4? You need to click the accept button on that page.
Command you're running looks fine, it will be smt like docker run ... -e TOKEN=awdwadhwa ...nicklucche/stable-diffusion, no spaces between TOKEN= and the actual token

[aeon3]

Oh no, I actually put the token in between these < >

How can I edit it?

I tried running again but it says container name "/stable-diffusion" is already in use.

[NickLucche]

you need to remove the created container first with docker rm stable-diffusion and then try the command docker run .. again.
Once you're able to run it with no errors the first time, you can then use docker start stable-diffusion to (more) quickly bring that up again. Let me know if that worked for you.

[aeon3]

ok it worked! thank you.

HOWEVER, I ran -e DEVICES=all ... but my hwmonitor shows 2nd gpu utilization as 5%, while the 1st one does the usual 90%.

What could be going on?

[NickLucche]

have you tried increasing the number of generated images? Multi-gpu works by splitting the workload evenly among devices. If that doesn't work, you might have an old image on your system, you can verify that by listing docker images | grep stable-diffusion

[aeon3]

have you tried increasing the number of generated images? Multi-gpu works by splitting the workload evenly among devices. If that doesn't work, you might have an old image on your system, you can verify that by listing docker images | grep stable-diffusion

this happens:
PS C:\Windows\system32> docker images | grep stable-diffusion
grep : The term 'grep' is not recognized as the name of a cmdlet, function, script file, or operable program. Check
the spelling of the name, or if a path was included, verify that the path is correct and try again.
At line:1 char:17

• docker images | grep stable-diffusion

•             ~~~~
  

  • CategoryInfo : ObjectNotFound: (grep:String) [], CommandNotFoundException
  • FullyQualifiedErrorId : CommandNotFoundException

[NickLucche]

are you executing the command inside wsl console? This is a bash command..
Have you also tried generating 2 images? That should get your second gpu to work as well.

My guess is that you're running an older version of the image, pre multi-gpu support, is that possible (1-2 weeks old)?
Regardless, if you could manage to report the boot-up logs of the container that would also be helpful