feat: package llama.cpp OpenCL backend for Qualcomm midstream - #2
Closed
chenchongbiao wants to merge 1 commit into
Closed
feat: package llama.cpp OpenCL backend for Qualcomm midstream#2chenchongbiao wants to merge 1 commit into
chenchongbiao wants to merge 1 commit into
Conversation
Build llama.cpp for Q6A and Q8B with the Adreno-optimized OpenCL backend. Cross-compile directly in the Debian 12 devcontainer with the GNU AArch64 toolchain so standard x86 GitHub runners do not need a nested Docker build. Keep GGML_NATIVE disabled, target OpenCL 2.0, embed the OpenCL kernels, and disable the Hexagon backend. Lock the source submodule to upstream tag b7992 and depend on qcom-adreno-opencl-midstream. Signed-off-by: Chongbiao Chen <chenchongbiao@radxa.com>
Contributor
|
No linked issues found. Please add the corresponding issues in the pull request description. |
Member
|
这个是怎么回事 |
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Build llama.cpp for Q6A and Q8B with the Adreno-optimized OpenCL backend.
Cross-compile directly in the Debian 12 devcontainer with the GNU AArch64 toolchain so standard x86 GitHub runners do not need a nested Docker build. Keep GGML_NATIVE disabled, target OpenCL 2.0, embed the OpenCL kernels, and disable the Hexagon backend.
Lock the source submodule to upstream tag b7992 and depend on qcom-adreno-opencl-midstream.