feat: apply upstream llama.cpp patch to fix Gemma-4 inference issues

2026-04-06 10:25:42 +01:00
parent bbd8f02014
commit d0281bb526
1 changed files with 1 additions and 0 deletions
@@ -62,6 +62,7 @@ RUN git clone -b ${BRANCH} --single-branch --recursive ${REPO} . \
 COPY llama-grammar.patch /tmp/llama-grammar.patch

 RUN patch -p1 < /tmp/llama-grammar.patch \
+  && curl -sSL https://github.com/ggml-org/llama.cpp/pull/21506.patch | patch -p1 \
  && cmake -S . -B build \
  -DGGML_HIP=ON \
  -DAMDGPU_TARGETS=gfx1151 \