NOTE: Update: 21/12/2024 so, tested automatic1111 and comfy UI with the updated packages rocm-rel-6.2.3 do note it has to manually remove the dependencies for the Nvidia version of py torch with comfy UI when installing the requirements, so that might be issues with some other ones, anyway, I tried getting the official image running but my computer couldn't even handle downloading it Because I think it tries extracting it then runs out of memory ( even though I have 64 GB, if you know a setting to stop it just gobbling the ram when extracting archives please do say ), but otherwise seems to work, testing flux in comfy UI will be a second
Update: 05/12/2024
looks like AMD is providing their own prebuilt images if you would like to use the those, however the pretty chunky https://hub.docker.com/r/rocm/pytorch/tags and they updated the rocm version for WSL (Link has docker instructions to) https://rocm.docs.amd.com/projects/radeon/en/latest/docs/install/wsl/install-radeon.html I need to update the version that's being used in the repo, and do some tests
haven't been able to get flux.dev working it always freezes up the entire computer when processing vae tell me if you manage to get past this for flux.dev
technically fixed it I think it's an issue with VAE steps in the driver, I recently upgraded to 64GB of ram from 32GB, and when the point it would usually crash when doing the VAE,
it is when it spikes up in memory and Does the same thing but doesn't crash,
remedy? if you don't have the hardware may be temporarily increasing your ram cash/whatever it's called file so it can temporarily allocate it, interestingly don't seem to spike after it's done it 1st/2nd time, Probably some JIT compiling on the AMD side using too much memory
docker is a bit weird inside of the virtual WSL2 Linux so you'll have to run a plane docker composed before it will allow the GPU pass through to work for some reason ( is part of the guide so don't worry too much )
it seems for me every time after a computer restart? I have to run the WSL2_set_up then exit out with ctrl + c then run the thing I want to run otherwise it complains about a missing bind
confirm working with Windows10 and an RX7900XTX
This is currently setup for automatic1111 , comfyUI , swarmUI (with comfyUI backend) and forge_ui,
good news, I found the source of the crash
it will crash when generating images , when on the final step workaround, turn off instant replay,
the initial model loading takes ages for some reason, haven't worked out how to cash it on automatic_1111
24.7.1 or newer AMD driver software version
an RDNA3 card 7XXX series, might work with 6XXX? idk Not able to test someone tried an RX 6950 XT and the GPU was being recognised in the container, no confirmation on if it worked properly tho
docker desktop (when installed go to settings gear icon > resources > wsl integration > Enable Integration with additional distro ( You might need to add one if it's not visible Just search for ubuntu in window search and it comes up and set up) )
WSL2
plenty of free storage space, maybe like 50gb? (for docker you can specify where the disc images located, settings gear icon > resources > advanced > disk image location)
a hefty amount of RAM since the build likes to consume a lot for some reason best set in your WSL config
C:\Users\<account name>\.wslconfig
( if not present add it, make sure to turn on file extensions in Windows Explorer so you don't get something like .wslconfig.txt )
[wsl2]
memory=50GB
swap=50GB
(Once you finish building your images It's advisable to turn memory back down to something that doesn't consume all your memory on your computer, since (docker) Linux does that thing where it thinks it's free memory so fills it up with random stuff but Windows doesn't know about that so you end up with your memory being filled up)
and then Restart wsl with this in Windows terminal wsl --shutdown then docker desktop should prompt to restart wsl and you allow it to do so
now, open a terminal in /wsl_2_rocm_docker_win and run docker file (if you haven't Built it already)
docker build -t wsl_2_rocm_docker_win .
wait for it to finish ( this may take a while )
and then,
go to /automatic_1111 and do
docker build -t automatic_1111 .
wait for it to finish ( this may take a while )
Load up a ubuntu wsl instance,
connect to it using the terminal , searching for ubuntu should bring up the terminal in Windows
and now, go to WSL2_set_up ( why? because for some reason if you don't run a normal docker then, run the one with the GPU Pass through it bugs out weirdly and says it can't find the GPU )
then run
( remembering you have set up the docker integration )
sudo docker compose up
wait for it to download and when it says it's running in docker desktop then press
ctrl+c to exit out of it (or close in docker desktop )
go to
and then navigate to the compose in automatic_1111 and do
sudo docker compose up
(it will download a few extra requirements since not quite sure how to pin them down)
and then it should load up,
you need to then get the internal IP of your wsl container
wsl --list --verbose
then pick out the name for the one you're running the Docker compose on
wsl -d <DistributionName> hostname -I
that should give you an IP then simply do http://<foundIP>:7860/
and then you should be connected to the UI
it will take a while since the model loading takes ages you can observe by your RAM slowly climbing, best to wait for it
then have fun!
now, open a terminal in /wsl_2_rocm_docker_win and run docker file (if you haven't Built it already)
docker build -t wsl_2_rocm_docker_win .
wait for it to finish ( this may take a while )
and now,
go to /comfy_ui and do
docker build -t comfy_ui .
wait for it to finish ( this may take a while )
Load up a ubuntu wsl instance,
connect to it using the terminal , searching for ubuntu should bring up the terminal in Windows
and now, go to WSL2_set_up ( why? because for some reason if you don't run a normal docker then, run the one with the GPU Pass through it bugs out weirdly and says it can't find the GPU )
then run
( remembering you have set up the docker integration )
sudo docker compose up
wait for it to download and when it says it's running in docker desktop then press
ctrl+c to exit out of it (or close in docker desktop )
go to
navigate to the docker compose in comfy_ui and do
sudo docker compose up and then it should load up,
(Of course then provide your own model in the models/checkpoints folder)
you need to then get the internal IP of your wsl container
wsl --list --verbose
then pick out the name for the one you're running the Docker compose on
wsl -d <DistributionName> hostname -I
that should give you an IP then simply do http://<foundIP>:8188/
and then you should be connected to the UI
if you got the model installed the and everything setup it should work! have fun
so, First off follow the comfy UI steps until Load up a ubuntu wsl instance, now go to /swarm_ui and do
docker build -t swarm_ui .
wait for it to finish ( this may take a while )
and now Load up a ubuntu wsl instance,
connect to it using the terminal , searching for ubuntu should bring up the terminal in Windows
and now, go to WSL2_set_up ( why? because for some reason if you don't run a normal docker then, run the one with the GPU Pass through it bugs out weirdly and says it can't find the GPU )
then run
then run
( remembering you have set up the docker integration )
sudo docker compose up
wait for it to download and when it says it's running in docker desktop then press
ctrl+c to exit out of it (or close in docker desktop )
go to
navigate to the docker compose in swarm_ui and do
sudo docker compose up and then it should load up,
(Of course then provide your own model in the models/checkpoints/OfficialStableDiffusion ( So it aligns nicely with what swarm UI Wants ) folder)
you need to then get the internal IP of your wsl container
wsl --list --verbose
then pick out the name for the one you're running the Docker compose on
wsl -d <DistributionName> hostname -I
that should give you an IP then simply do http://<foundIP>:7801/
and then you should be connected to the UI
if you got the model installed the and everything setup it should work! have fun
now, open a terminal in /wsl_2_rocm_docker_win and run docker file (if you haven't Built it already)
docker build -t wsl_2_rocm_docker_win .
wait for it to finish ( this may take a while )
and then,
go to /forge_ui and do
docker build -t forge_ui .
wait for it to finish ( this may take a while )
Load up a ubuntu wsl instance,
connect to it using the terminal , searching for ubuntu should bring up the terminal in Windows
and now, go to WSL2_set_up ( why? because for some reason if you don't run a normal docker then, run the one with the GPU Pass through it bugs out weirdly and says it can't find the GPU )
then run
( remembering you have set up the docker integration )
sudo docker compose up
wait for it to download and when it says it's running in docker desktop then press
ctrl+c to exit out of it (or close in docker desktop )
go to
and then navigate to the compose in forge_ui and do
sudo docker compose up
(it will download a few extra requirements since not quite sure how to pin them down)
and then it should load up,
you need to then get the internal IP of your wsl container
wsl --list --verbose
then pick out the name for the one you're running the Docker compose on
wsl -d <DistributionName> hostname -I
that should give you an IP then simply do http://<foundIP>:7860/
and then you should be connected to the UI
then have fun!