feat: allow building Docker image without vllm

- Simplify build process by bundling ppa key
- Add new BUILD_WITH_VLLM build-arg
- Add `make local_build` option
- Build ffmpeg in the builder image
- Build ffmpeg using all available cores
- Remove no longer needed dependencies

Dockerfile

@@ -1,23 +1,22 @@
 ARG BASE_IMAGE_BUILD=nvidia/cuda:12.2.2-cudnn8-devel-ubuntu22.04
 ARG BASE_IMAGE_RUN=nvidia/cuda:12.2.2-cudnn8-runtime-ubuntu22.04
+ARG BUILD_WITH_VLLM=1
 
 ## Base Image
 
 FROM ${BASE_IMAGE_BUILD} AS builder
 
 RUN \
     apt-get update && \
-    apt-get install -y apt-transport-https ca-certificates gnupg git
+    apt-get install -y apt-transport-https ca-certificates
 
 COPY docker/rootfs/ /
 
 RUN \
-    apt-dpkg-wrap apt-key adv --keyserver hkp://keyserver.ubuntu.com:80 --recv-keys F23C5A6CF475977595C89F51BA6932366A755776 && \
     apt-dpkg-wrap apt-get update && \
-    apt-dpkg-wrap apt-get install -y build-essential libcurl4-openssl-dev python3.11 python3.11-venv && \
-    apt-cleanup
+    apt-dpkg-wrap apt-get install -y build-essential python3.11 python3.11-venv
 
-COPY requirements-vllm.txt /app/
+COPY requirements*.txt /app/
 
 WORKDIR /app
 
@@ -26,23 +25,23 @@ ENV PIP_DISABLE_PIP_VERSION_CHECK=on
 RUN \
     python3.11 -m venv .venv && \
     . .venv/bin/activate && \
-    pip install -vvv -r requirements-vllm.txt
+    if [ "${BUILD_WITH_VLLM}" = "1" ]; then \
+        echo "Building with VLLM"; \
+        pip install -vvv -r requirements-vllm.txt; \
+    else \
+        echo "Building without VLLM"; \
+        pip install -vvv -r requirements.txt; \
+    fi
 
 ## Build ffmpeg
 
-FROM ${BASE_IMAGE_RUN} AS ffmpeg_install
-
-COPY docker/rootfs/ /
+FROM builder AS ffmpeg_builder
 
 # ffmpeg build dependencies
 RUN \
-    apt-dpkg-wrap apt-key adv --keyserver hkp://keyserver.ubuntu.com:80 --recv-keys F23C5A6CF475977595C89F51BA6932366A755776 && \
-    apt-dpkg-wrap apt-get update && \
     apt-dpkg-wrap apt-get install -y \
         autoconf \
         automake \
-        build-essential \
-        cmake \
         libopus-dev \
         libopus0 \
         libtool \
@@ -51,8 +50,7 @@ RUN \
         wget \
         yasm \
         zlib1g \
-        zlib1g-dev && \
-    apt-cleanup
+        zlib1g-dev
 
 # Build ffmpeg6 (required for pytorch which only supports ffmpeg < v7)
 RUN \
@@ -66,52 +64,38 @@ RUN \
       --enable-shared \
       --enable-gpl \
       --enable-libopus && \
-    make && \
+    JOBS="$(nproc)" && \
+    make -j "${JOBS}" && \
     make install && \
     ldconfig
 
-RUN \
-    apt-dpkg-wrap apt-get autoremove -y \
-        autoconf \
-        automake \
-        build-essential \
-        cmake \
-        libopus-dev \
-        libtool \
-        pkg-config \
-        texinfo \
-        wget \
-        yasm \
-        zlib1g-dev
-
 ## Production Image
 
-FROM ffmpeg_install
+FROM ${BASE_IMAGE_RUN}
+
+COPY --chown=jitsi:jitsi docker/run-skynet.sh /opt/
+COPY --from=ffmpeg_builder /usr/local/include /usr/local/include
+COPY --from=ffmpeg_builder /usr/local/lib/lib* /usr/local/lib/
+COPY --from=ffmpeg_builder /usr/local/lib/pkgconfig /usr/local/lib/pkgconfig
+COPY --chown=jitsi:jitsi --from=builder /app/.venv /app/.venv
+COPY --chown=jitsi:jitsi /skynet /app/skynet/
 
 RUN \
     apt-get update && \
-    apt-get install -y apt-transport-https ca-certificates gnupg
+    apt-get install -y apt-transport-https ca-certificates
 
 COPY docker/rootfs/ /
-COPY --chown=jitsi:jitsi docker/run-skynet.sh /opt/
 
 RUN \
-    apt-dpkg-wrap apt-key adv --keyserver hkp://keyserver.ubuntu.com:80 --recv-keys F23C5A6CF475977595C89F51BA6932366A755776 && \
-    apt-dpkg-wrap apt-get update && \
-    apt-dpkg-wrap apt-get install -y python3.11 python3.11-venv tini libgomp1 strace gdb && \
+    apt-get update && \
+    apt-dpkg-wrap apt-get install -y python3.11 python3.11-venv tini libgomp1 libopus0 zlib1g strace gdb && \
     apt-cleanup
 
 # Principle of least privilege: create a new user for running the application
 RUN \
     groupadd -g 1001 jitsi && \
     useradd -r -u 1001 -g jitsi jitsi
 
-# Copy virtual environment
-COPY --chown=jitsi:jitsi --from=builder /app/.venv /app/.venv
-
-# Copy application files
-COPY --chown=jitsi:jitsi /skynet /app/skynet/
-
 ENV \
     # https://docs.python.org/3/using/cmdline.html#envvar-PYTHONUNBUFFERED
     PYTHONUNBUFFERED=1 \
@@ -120,8 +104,7 @@ ENV \
     PYTHONPATH=/app \
     OUTLINES_CACHE_DIR=/app/vllm/outlines \
     VLLM_CONFIG_ROOT=/app/vllm/config \
-    HF_HOME=/app/hf  \
-    LLAMA_PATH="/models/Llama-3.1-8B-Instruct-Q8_0.gguf"
+    HF_HOME=/app/hf
 
 VOLUME [ "/models" ]
 

Makefile

@@ -10,10 +10,21 @@ _login:
 
 build : _login
 	docker buildx build \
-	--progress plain \
-	--push \
-	--platform ${PLATFORMS} \
-	-t ${IMAGE_REGISTRY}/skynet:summaries-${GIT_HASH} \
-	-t ${IMAGE_REGISTRY}/skynet:whisper-${GIT_HASH} \
-	-t ${IMAGE_REGISTRY}/skynet:${GIT_HASH} \
-	-t ${IMAGE_REGISTRY}/skynet:latest .
+		--progress plain \
+		--push \
+		--platform ${PLATFORMS} \
+		-t ${IMAGE_REGISTRY}/skynet:summaries-${GIT_HASH} \
+		-t ${IMAGE_REGISTRY}/skynet:whisper-${GIT_HASH} \
+		-t ${IMAGE_REGISTRY}/skynet:${GIT_HASH} \
+		-t ${IMAGE_REGISTRY}/skynet:latest \
+		.
+
+local_build:
+	docker buildx build \
+		--progress plain \
+		--load \
+		--build-arg BASE_IMAGE_BUILD=ubuntu:22.04 \
+		--build-arg BASE_IMAGE_RUN=ubuntu:22.04 \
+		--build-arg BUILD_WITH_VLLM=0 \
+		-t jitsi/skynet:latest \
+		.

docker/rootfs/etc/apt/sources.list.d/python-ppa.list

@@ -1 +1 @@
-deb https://ppa.launchpadcontent.net/deadsnakes/ppa/ubuntu jammy main
+deb [signed-by=/opt/python-ppa.gpg] https://ppa.launchpadcontent.net/deadsnakes/ppa/ubuntu jammy main