mirror of
https://github.com/huggingface/text-generation-inference.git
synced 2025-10-09 15:05:24 +00:00
49 lines
1.4 KiB
Docker
49 lines
1.4 KiB
Docker
ARG CUDA_ARCH_LIST="75-real;80-real;86-real;89-real;90-real"
|
|
|
|
# Build dependencies resolver stage
|
|
FROM lukemathwalker/cargo-chef:latest as chef
|
|
WORKDIR /usr/src/text-generation-inference
|
|
|
|
FROM chef as planner
|
|
COPY . .
|
|
RUN cargo chef prepare --recipe-path recipe.json
|
|
|
|
# CUDA dependent dependencies resolver stage
|
|
FROM nvcr.io/nvidia/pytorch:24.05-py3 as cuda-builder
|
|
|
|
RUN apt update && apt install -y \
|
|
cmake \
|
|
gcc \
|
|
g++ \
|
|
git \
|
|
git-lfs \
|
|
ninja-build
|
|
|
|
# Install TensorRT
|
|
COPY backends/trtllm/scripts/install_tensorrt.sh /opt/install_tensorrt.sh
|
|
RUN chmod +x /opt/install_tensorrt.sh && \
|
|
/opt/install_tensorrt.sh
|
|
|
|
# Install Rust
|
|
RUN curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | bash -s -- -y && \
|
|
chmod -R a+w $HOME/.rustup && \
|
|
chmod -R a+w $HOME/.cargo
|
|
|
|
ENV PATH="$HOME/.cargo/bin:$PATH"
|
|
RUN $HOME/.cargo/bin/cargo install cargo-chef
|
|
|
|
# Backend build step
|
|
WORKDIR /usr/src/text-generation-inference
|
|
|
|
# Cache dependencies
|
|
COPY --from=planner /usr/src/text-generation-inference/recipe.json .
|
|
RUN $HOME/.cargo/bin/cargo chef cook --release --recipe-path recipe.json
|
|
|
|
# Build actual TGI
|
|
COPY . .
|
|
RUN $HOME/.cargo/bin/cargo build --release --bin text-generation-backends-trtllm
|
|
|
|
FROM nvcr.io/nvidia/pytorch:24.05-py3
|
|
WORKDIR /opt
|
|
COPY --from=cuda-builder /usr/local/tensorrt /usr/local/tensorrt
|
|
COPY --from=cuda-builder /usr/src/text-generation-inference/target/release/text-generation-backends-trtllm /opt/text-generation-launcher |