huggingface/text-generation-inference
5
0
Fork 0
mirror of https://github.com/huggingface/text-generation-inference.git synced 2025-05-04 16:52:06 +00:00
Code Issues Packages Projects Releases Wiki Activity

fix: update trtllm dockefile after rebase

Browse Source
This commit is contained in:
drbh 2024-12-13 16:59:19 +00:00
parent bb00fb33ba
commit 91ed362e74
1 changed files with 21 additions and 16 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

35
Dockerfile_trtllm
View File

@ -1,5 +1,5 @@
ARG CUDA_ARCH_LIST="75-real;80-real;86-real;89-real;90-real" ARG CUDA_ARCH_LIST="75-real;80-real;86-real;89-real;90-real"
ARG OMPI_VERSION="4.1.6" ARG OMPI_VERSION="4.1.7rc1"
# Build dependencies resolver stage # Build dependencies resolver stage
FROM lukemathwalker/cargo-chef:latest AS chef FROM lukemathwalker/cargo-chef:latest AS chef
@ -10,7 +10,7 @@ COPY . .
RUN cargo chef prepare --recipe-path recipe.json RUN cargo chef prepare --recipe-path recipe.json
# CUDA dependent dependencies resolver stage # CUDA dependent dependencies resolver stage
FROM nvidia/cuda:12.6.1-cudnn-devel-ubuntu22.04 AS cuda-builder FROM nvidia/cuda:12.6.3-cudnn-devel-ubuntu24.04 AS cuda-builder
RUN --mount=type=cache,target=/var/cache/apt,sharing=locked \ RUN --mount=type=cache,target=/var/cache/apt,sharing=locked \
--mount=type=cache,target=/var/lib/apt,sharing=locked \ --mount=type=cache,target=/var/lib/apt,sharing=locked \
@ -18,18 +18,13 @@ RUN --mount=type=cache,target=/var/cache/apt,sharing=locked \
build-essential \ build-essential \
cmake \ cmake \
curl \ curl \
gcc \ gcc-14 \
g++ \ g++-14 \
git \ git \
git-lfs \ git-lfs \
libssl-dev \ libssl-dev \
libucx-dev \
ninja-build \ ninja-build \
pkg-config \
python3 \
python3-dev \
python3-setuptools \
tar \
wget \
ffmpeg \ ffmpeg \
libavcodec-dev \ libavcodec-dev \
libavfilter-dev \ libavfilter-dev \
@ -38,8 +33,13 @@ RUN --mount=type=cache,target=/var/cache/apt,sharing=locked \
libavutil-dev \ libavutil-dev \
libswscale-dev \ libswscale-dev \
pkg-config \ pkg-config \
libclang-dev \ pipx \
clang python3 \
python3-dev \
python3-setuptools \
tar \
wget && \
pipx ensurepath
ENV TGI_INSTALL_PREFIX=/usr/local/tgi ENV TGI_INSTALL_PREFIX=/usr/local/tgi
ENV TENSORRT_INSTALL_PREFIX=/usr/local/tensorrt ENV TENSORRT_INSTALL_PREFIX=/usr/local/tensorrt
@ -93,13 +93,15 @@ RUN mkdir $TGI_INSTALL_PREFIX && mkdir "$TGI_INSTALL_PREFIX/include" && mkdir "$
cd backends/trtllm && \ cd backends/trtllm && \
CMAKE_INSTALL_PREFIX=$TGI_INSTALL_PREFIX cargo build --release CMAKE_INSTALL_PREFIX=$TGI_INSTALL_PREFIX cargo build --release
FROM nvidia/cuda:12.6.1-cudnn-runtime-ubuntu22.04 AS runtime FROM nvidia/cuda:12.6.3-cudnn-runtime-ubuntu24.04 AS runtime
RUN apt update && apt install -y python3-minimal python3-dev python3-pip && \ RUN apt update && apt install -y libucx0 pipx python3-minimal python3-dev python3-pip python3-venv && \
rm -rf /var/lib/{apt,dpkg,cache,log}/ && \ rm -rf /var/lib/{apt,dpkg,cache,log}/ && \
python3 -m pip install transformers tokenizers pipx ensurepath && \
pipx install --include-deps transformers tokenizers
WORKDIR /usr/local/tgi/bin WORKDIR /usr/local/tgi/bin
ENV PATH=/root/.local/share/pipx/venvs/transformers/bin/:$PATH
ENV LD_LIBRARY_PATH="/usr/local/tgi/lib:/usr/local/mpi/lib:/usr/local/tensorrt/lib:/usr/local/cuda/lib64/stubs:$LD_LIBRARY_PATH" ENV LD_LIBRARY_PATH="/usr/local/tgi/lib:/usr/local/mpi/lib:/usr/local/tensorrt/lib:/usr/local/cuda/lib64/stubs:$LD_LIBRARY_PATH"
ENV TOKENIZERS_PARALLELISM=false ENV TOKENIZERS_PARALLELISM=false
ENV OMPI_MCA_plm_rsh_agent="" ENV OMPI_MCA_plm_rsh_agent=""
@ -108,6 +110,9 @@ COPY --from=mpi-builder /usr/local/mpi /usr/local/mpi
COPY --from=trt-builder /usr/local/tensorrt /usr/local/tensorrt COPY --from=trt-builder /usr/local/tensorrt /usr/local/tensorrt
COPY --from=tgi-builder /usr/local/tgi /usr/local/tgi COPY --from=tgi-builder /usr/local/tgi /usr/local/tgi
COPY --from=tgi-builder /usr/src/text-generation-inference/target/release/text-generation-backends-trtllm /usr/local/tgi/bin/text-generation-launcher COPY --from=tgi-builder /usr/src/text-generation-inference/target/release/text-generation-backends-trtllm /usr/local/tgi/bin/text-generation-launcher
# Copy the ffmpeg libraries
COPY --from=cuda-builder /usr/lib/x86_64-linux-gnu/* /usr/lib/x86_64-linux-gnu-copy/
ENV LD_LIBRARY_PATH="$LD_LIBRARY_PATH:/usr/lib/x86_64-linux-gnu-copy"
FROM runtime FROM runtime