fix: update image to support usage info

.github/workflows/build-publish.yaml

@@ -53,6 +53,13 @@ jobs:
           sudo apt-get update
           sudo apt-get install -y qemu-user-static
 
+      - name: Clone llama-cpp-python
+        run: |
+          git config --global user.email "[email protected]"
+          git config --global user.name "Podman Desktop team"
+          cd chat
+          ./setup.sh
+
       - name: Build Playground Image
         id: build-image
         uses: redhat-actions/buildah-build@v2

.github/workflows/pr-check.yaml

@@ -49,6 +49,13 @@ jobs:
           sudo apt-get update
           sudo apt-get install -y qemu-user-static
 
+      - name: Clone llama-cpp-python
+        run: |
+          git config --global user.email "[email protected]"
+          git config --global user.name "Podman Desktop team"
+          cd chat
+          ./setup.sh
+
       - name: Build Playground Image
         id: build-image
         uses: redhat-actions/buildah-build@v2

.github/workflows/release.yaml

@@ -89,6 +89,13 @@ jobs:
           sudo apt-get update
           sudo apt-get install -y qemu-user-static
 
+      - name: Clone llama-cpp-python
+        run: |
+          git config --global user.email "[email protected]"
+          git config --global user.name "Podman Desktop team"
+          cd chat
+          ./setup.sh
+
       - name: Build Playground Image
         id: build-image
         uses: redhat-actions/buildah-build@v2
@@ -130,4 +137,4 @@ jobs:
         with:
           tag: ${{ needs.tag.outputs.githubTag }}
           name: ${{ needs.tag.outputs.githubTag }}
-
+

.gitignore

@@ -0,0 +1 @@
+llama-cpp-python

chat/base/Containerfile

@@ -21,6 +21,7 @@ RUN microdnf install -y python3 python3-pip gcc g++ shadow-utils && microdnf cle
 RUN useradd -r -g root -m -d /home/default -s /bin/bash default
 WORKDIR /home/default
 COPY requirements.txt requirements.txt
+COPY llama-cpp-python llama-cpp-python
 RUN pip install --no-cache-dir --upgrade -r requirements.txt
 USER default
 COPY run.sh run.sh

chat/cuda/amd64/Containerfile

@@ -20,6 +20,7 @@ RUN  dnf install -y gcc-toolset-13-gcc gcc-toolset-13-gcc-c++
 USER 1001
 WORKDIR /locallm
 COPY requirements.txt .
+COPY --chown=1001 llama-cpp-python llama-cpp-python
 COPY run.sh .
 ENV CMAKE_ARGS="-DGGML_CUDA=on -DLLAMA_AVX2=OFF -DLLAMA_FMA=OFF -DLLAMA_F16C=OFF"
 ENV FORCE_CMAKE=1

chat/requirements.txt

@@ -1,2 +1,2 @@
-llama-cpp-python[server]==0.3.2
+./llama-cpp-python[server]
 pip==24.3.1

chat/setup.sh

@@ -0,0 +1,6 @@
+git clone https://github.com/abetlen/llama-cpp-python --no-checkout --recurse-submodules
+cd llama-cpp-python
+git checkout v0.3.2
+git submodule update
+git fetch origin pull/1552/head
+git rebase ffc47e5a95cf31fe6d9d6d7953d5eb17fdcc8513

chat/vulkan/amd64/Containerfile

@@ -32,6 +32,7 @@ RUN dnf install -y python3-dnf-plugin-versionlock && \
 
 WORKDIR /locallm
 COPY requirements.txt ./
+COPY llama-cpp-python llama-cpp-python
 RUN pip install --upgrade pip
 ENV CMAKE_ARGS="-DGGML_VULKAN=on"
 ENV FORCE_CMAKE=1

chat/vulkan/arm64/Containerfile

@@ -30,6 +30,7 @@ https://dl.fedoraproject.org/pub/epel/epel-release-latest-9.noarch.rpm && \
     dnf install -y glslc && \
     dnf clean all
 COPY requirements.txt ./
+COPY llama-cpp-python llama-cpp-python
 RUN pip install --upgrade pip
 ENV CMAKE_ARGS="-DGGML_VULKAN=on"
 ENV FORCE_CMAKE=1