mickr777

Is there a reason your want to edit that file? if its to change config it should be in the userfiles/invokeai.yaml file and permissions changed in the file manager in unraid ui to read/write (btw this docker is not the same as the one in the community apps)

Base image updated to Ubuntu 24.04 for Python 3.12 To avoid errors, you’ll need to manually delete the following folders before restarting the docker: /mnt/cache/appdata/invokeai/venv /mnt/cache/appdata/invokeai/invokeai Do not delete your /mnt/cache/appdata/invokeai/userfiles folder — this contains your model, database, output images Once removed, restart the docker and it will automatically clone the git and rebuild the venv and frontend files

Max vram cache is how much vram invoke can use to cache models in for repeated uses. Eg not off load them to ram and keep in vram. Setting it too high doesn't leave room for invoke to use when working and can cause OOM errors. 60-70% of your vram for cache is what I found ideal, the 100mb was just to disable it and see if that was the cause Working ram does nearly the same thing, but it tells invoke how much vram you want to keep free to use when working. I have set my to 12gb as I use some node that don't use the caching system and avoids OOM when there used

its related to the new memory caching changes in invoke, https://invoke-ai.github.io/InvokeAI/features/low-vram/ Try adding max_cache_vram_gb:0.1 to your invokeai.yaml file under userfiles in the dockers appdata folder for a test (this disables vram caching of models) see if that helps, if it does try some of the other options in the above link. you probably dont need to go this far with 24gb vram, but I can load and use the full flux dev model and full text encoder with 12gb vram now, on my laptop using: attention_type: torch-sdp enable_partial_loading: true force_tiled_decode: true keep_ram_copy_of_weights: false device_working_mem_gb: 12 there is a trade off with partial loading and tiled decode, as does slow it down a little and if you have loads of ram dont worry about keep ram copy false,

sadly no, Invokeai it self doesn't support multiple gpus

If anyone is still using this docker besides me, since there is official unraid docker in the community app plugin. the latest update i just pushed to this docker will require you to go back to port 9090, (this is to avoid a slow down in the dev ui on port 5173)

Ok made it for the training ui, port 1234 is the training ui, port 2345 is the tensorboard ui to watch progress of training my-Invoke-training.xml

So am I maybe just as test set it back to a empty userfiles under appdata, see if it still does it

that should be fine as my userfiles are on the array, not in appdata

Ok, try replacing the invokeai.yaml under userfiles with this file invokeai.yaml

double check all folders under the invokeai appdata folder are set to owner "nobody"

should still be able to add as a variable HUGGING_FACE_HUB_TOKEN

Trying to replicate this error, is this a fresh install? or after an update to to and install?

Still active, the github repo got changed a while back, you can update that in you xml file (its not in the community apps if that where you were looking), all info in first post should be still up to date

Just did a fresh install and i couldnt get it to give that error, I repushed docker image to docker hub see if that helps

The empty folders should be made by the xml for them to be in the appdata eg mine is my-invokeai.xml (bit like an external symbolic link), then the script can use them like any other folder, but the data doesnt get wiped on docker image changes

I ended up having to have anything that I wanted in to the folders in appdata to be installed or ran in the start.sh script, if you want more info on that you can see all that in the main post under full manual install (I left comments in start.sh code to show what each section of it does), so in realty the docker image just has all the basic prerequisites to get the main install done with the start.sh. this was my first ever docker, so i just winged it and hoped for the best and it worked for unraid 😁

Updated onnxruntime and xformers

Updated to Support the change from yarn classic to pnpm

Yes but not easily, but my issue is I dont have a amd card to test it on

what docker image are you using? as models config should be here /home/invokeuser/userfiles/configs/models.yaml not where it is looking for it in /root/invokeai/configs/models.yaml with my docker image in first post

there was a issue where the needed models were not downloaded, to force a download, delete the file appdata/invokeai/venv/pyvenv.cfg only in unraid and rerun the docker, this will force the download of missing models.

fix for UI not loading I forced the docker to use the dev WebUI. if you are using the Dockerhub file: click the invokeai container and click update once updated right click the container and go edit change this line to the following next find webui port and click edit and change it to the following to access the webui now you have to use yourunraidip:5173. if you used the manual install: I have updated the files in main post you will need to rebuild your image and container with them. once built change the ports as the same as above also if you get startup aborted: there was a issue where the needed models were not downloaded, to force a download, delete the file /invokeai/venv/pyvenv.cfg and rerun the docker, this will force the download of missing models.

https://invoke-ai.github.io/InvokeAI/features/PROMPTS/#how-dynamic-prompts-work That might help If you have discord you should join the invokeai server https://discord.gg/ZmtBAhwWhy Also if it using the gpu you should see it on start in your docker logs

If you on the manual setup Should only have to delete and rebuild the container with the new files, if your on the docker hub images remove container delete image, then add docker back but change the repo to mickr777/invokeai_unraid_main and start it should migrate everything for you 🤞(minus your outputs folder it need to be backup, as it will not show in new version)