sh: 1: accelerate: not found
today im getting this error when using Kohya to caption images or trying to train a LoRA using a Stable Diffusion Kohya_ss ComfyUI Ultimate Pod. I have been using these pods for the last month without issue, but today nothing is working. My work flow has not change. I was successfully creating 2 days ago. Any ideas?
Thanks sh: 1: accelerate: not found
and /bin/sh: 1: accelerate: not found
Thanks sh: 1: accelerate: not found
and /bin/sh: 1: accelerate: not found
23 Replies
Do u have a pod id?
Maybe share it. Also I assume this replicates even when creating a new pod?
I can’t say as Im just another member, but hopefully when tmmrw PST wraps around u can ping a staff and they can look at ur pod id and logs
Ive tried 3 pods, and each time it takes about 30mins for the new pod to be ready.. I just closed my last pod 20secs ago
same on all pods, 3 different setups/builds
Yeah, I was thinking I must be in a different time zone and everyone is alseep
Sg! Whether u have a pod id, or just share the template link 🔗 i think is good :). Or steps u did to replicate
Like:
Commands u ran
Will be helpful to flag this to them - otherwise they cant really say. Or is this just completely happening in the logs and stopping it from ever starting up?
If the latter prob just can share the template name
Ah this looks like ashelyks pod
https://hub.docker.com/r/ashleykza/stable-diffusion-webui/tags
the Template is
Stable Diffusion Kohya_ss ComfyUI Ultimate
its happenig while using Kohya_SS to either captions images or train a LoRA. it fails to start the process and just ends.
My workflow and set up hasnt changed and i was succesfull 2 days ago on the same type of Pods.
My workflow and set up hasnt changed and i was succesfull 2 days ago on the same type of Pods.
Why dont u default back down to 3.11.0
I think check the template for the docker image + tag and deploy ur own to unblock for now
@ashleyk - just fyi - maybe something wrong with ur 3.11.1
It makes sense cause looks like there is a minor version bump 9 hours ago, maybe caused an issue
Im not very tech savvy that way.. I figgured most stuff out by following youtube videos. This is all new to me this month.
I see.
hopefully u can see this name? U can make ur own custom template
with the same settings
before u deploy do edit settings or something
screenshot it
then u can make ur own custom template
and in the docker image place just put the same thing but do
name…:3.11.0
9 hours ago there was a push which may be causing ur issue
Here in the container image just copy the name into it and u can set the storage space / ports / all that to the screenshot u take before u deploy the original template
Like this?
Mad ea new template, Im starting the pod up now.. Will see how it goes
I think you 100% correct about the Push update 9 hrs ago.
Thanks for you help
If your still around.. the Pod is up but when trying to log into Jupyter, the password isnt set up its asking for a Token How do i find this?
GitHub
GitHub - ashleykleynhans/stable-diffusion-docker: Docker image for ...
Docker image for Stable Diffusion WebUI with ControlNet, After Detailer, Dreambooth, Deforum and roop extensions, as well as Kohya_ss and ComfyUI - GitHub - ashleykleynhans/stable-diffusion-docker:...
Can try this!
The jup1ter thingy
is prob the password
thats what it normaly is but isnt working for this template -apprently
Jup1t3R!
U can try:
to open a webssh terminal and do jupyter server list
and see if u get a token through there
GitHub
GitHub - justinwlin/FooocusRunpod
Contribute to justinwlin/FooocusRunpod development by creating an account on GitHub.
ex on mine
okay. everything is working, so far all good.. Thanks
Glad was able to figure out a work around!
And from now on u can guarantee at least a stable version 🙂
if u want to get more technical 😂 u can download docker, pull down ashelyk’s image, retag it to ur username/imagename:tag, and push to ur own repo
can ask chatgpt for details
ull guarantee always having a copy of the image
for ur own stability
but for now this also good haha
It def is bit daunting at first maybe another’s day adventure lol
Apologies, tried to fix an issue with syncing to network volumes and I hard-coded the
fix_venv.sh script to fix venvs residing in /workspace/venv by accident and this caused issues with Kohya_ss because its venv is in /workspace/kohya_ss/venv and not /workspace/venv. Its been fixed in 3.11.2.