[AMD] Add AMD ROCm Docker support (RDNA3/RDNA4)#1292
Open
Kaihui-AMD wants to merge 1 commit into
Open
Conversation
Kaihui-AMD
force-pushed
the
add-rocm-docker-support
branch
from
700a38c to
fea78b2
Compare
Author
|
Hi @leng-yue, adds ROCm Docker support, verified on RDNA4 (gfx1201). Review when you have a chance, thanks! |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Feature
Fish Speech runs on AMD GPUs via ROCm. The ROCm image is based on the official
rocm/pytorchimage, which already ships a gfx-tuned PyTorch, so no separate torch install is needed. Verified on RDNA4 (Radeon AI PRO R9700 / gfx1201) with ROCm 7.2.3; RDNA3 (gfx1100/gfx1101) should also work.docker/Dockerfile.rocmbased onrocm/pytorch:rocm7.2.3(webui + server targets)compose.rocm.ymlwith/dev/kfd+/dev/dripassthroughdocs/en/install.mdThe ROCm base image ships a gfx-tuned PyTorch, so deps install without torch (--no-deps) to avoid pulling a CUDA build. pyaudio uses --no-build-isolation, and protobuf is pinned after install to satisfy the generated proto code.
Verified end-to-end on Radeon AI PRO R9700 (gfx1201/RDNA4) with ROCm 7.2.3, PyTorch 2.9.1: s2-pro inference + WebUI both working (bf16, AOTriton SDPA).
Is this PR adding new feature or fix a BUG?
Add feature: AMD ROCm Docker support (RDNA3/RDNA4), alongside the existing CUDA/CPU paths.
Is this pull request related to any issue? If yes, please link the issue.
No linked issue
Test