mirror/LocalAI
1
0
Fork 0
mirror of https://github.com/mudler/LocalAI.git synced 2026-04-29 19:44:13 -04:00
Code Issues Packages Projects Releases Wiki Activity

chore(whisperx): drop ROCm/hipblas build target (#9474)

Browse Source 
whisperx has no upstream AMD GPU support and its core transcription path
(faster-whisper -> ctranslate2) falls back to CPU on AMD since the PyPI
ctranslate2 is CUDA-only. The torch rocm wheels would accelerate only the
alignment/diarization stages, producing a misleadingly half-working image.

Drop the hipblas variant rather than shipping a partially accelerated build
users can't distinguish from the real thing. AMD hosts now fall through
the capability map to cpu-whisperx / cpu-whisperx-development.

Also removes the now-dangling rocm-whisperx assertion from
pkg/system/capabilities_test.go and the ROCm mention from the whisperx
row in docs/content/reference/compatibility-table.md.

Assisted-by: Claude Code:claude-opus-4-7
This commit is contained in:
Ettore Di Giacinto
2026-04-21 21:50:18 +02:00
committed by GitHub
parent a7dbb2a83d
commit 39573ecd2a
5 changed files with 1 additions and 33 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
backend/python/whisperx/requirements-hipblas.txt
View File

@@ -1,6 +0,0 @@
# whisperx hard-pins torch~=2.8.0, which is not available in the rocm7.x indexes
# (they start at torch 2.10). Keep rocm6.4 wheels here — they still load against
# the rocm7.2.1 runtime via AMD's forward-compatibility window.
--extra-index-url https://download.pytorch.org/whl/rocm6.4
torch==2.8.0+rocm6.4
whisperx @ git+https://github.com/m-bain/whisperX.git