Skip to content

fix: respect explicit QMD_LLAMA_GPU backends#531

Open
kuishou68 wants to merge 1 commit intotobi:mainfrom
kuishou68:fix-respect-llama-gpu-override
Open

fix: respect explicit QMD_LLAMA_GPU backends#531
kuishou68 wants to merge 1 commit intotobi:mainfrom
kuishou68:fix-respect-llama-gpu-override

Conversation

@kuishou68
Copy link
Copy Markdown
Contributor

Summary

  • preserve explicit QMD_LLAMA_GPU backends like metal, cuda, and vulkan
  • keep CPU-only override values mapped to false
  • add unit coverage for the GPU mode parser

Closes #529.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

QMD_LLAMA_GPU=metal is ignored when GPU mode falls back to "auto"

2 participants