Add documentation missing information for AMD devices.#8546
Add documentation missing information for AMD devices.#8546ArKam wants to merge 6 commits intoinvoke-ai:mainfrom
Conversation
Add information required to be able to run invoke on AMD devices.
ebr
left a comment
There was a problem hiding this comment.
Thank you for the contribution!
LGTM Co-authored-by: Eugene Brodsky <ebr@users.noreply.github.com>
|
Anyone that could merge this documentation update? |
|
@hipsterusername or @psychedelicious could you guys have a look at it? There are more and more issues created about ROCm support, and as I'm using it day to day I know the docker documentation is fine and working as I'm on AMD (fully) using up to ROCm 7.1 with invoke-ai with a GFX1030 (RX6600) and a GFX1201 (RX9070XT). |
|
@lstein and @blessedcoolant are the new maintainers you’ll want to have look at this. 🫡 |
Can you please share how you got it working? |
I just started working on the same thing! Id love to know of any progress. |
I tried it a few times with different strategies the last week and it always failed for different reasons. Did you get it working? Someone else seems to also work on updating rocm. |
Same! my biggest issue has been with the bits and bytes lib. Looks like better Strix Halo support is on its way but I could really use it now! |
|
bitsandbytes just released an update that might help! https://github.com/bitsandbytes-foundation/bitsandbytes/releases/tag/0.49.0 |
|
Well... Close but no cigar! I get the error below when trying to run SD1.x and SDXL models |
Add information required to be able to run invoke on AMD devices.
Summary
Improve invoke docker based installation documentation on AMD devices.
Related Issues / Discussions
None
QA Instructions
A simple straightforward documentation for people looking to run invoke on AMD based hardware.
Avoid users having to dig github issues and source code in order to find out docker requirements for AMD devices.
Merge Plan
None
Checklist