llama.cpp/.devops/main-rocm.Dockerfile

46 lines
1008 B
Docker
Raw Normal View History

ARG UBUNTU_VERSION=22.04
# This needs to generally match the container host's environment.
ARG ROCM_VERSION=5.6
# Target the CUDA build image
ARG BASE_ROCM_DEV_CONTAINER=rocm/dev-ubuntu-${UBUNTU_VERSION}:${ROCM_VERSION}-complete
FROM ${BASE_ROCM_DEV_CONTAINER} as build
# Unless otherwise specified, we make a fat build.
# List from https://github.com/ggerganov/llama.cpp/pull/1087#issuecomment-1682807878
# This is mostly tied to rocBLAS supported archs.
ARG ROCM_DOCKER_ARCH=\
gfx803 \
gfx900 \
gfx906 \
gfx908 \
gfx90a \
gfx1010 \
gfx1030 \
gfx1100 \
gfx1101 \
gfx1102
COPY requirements.txt requirements.txt
COPY requirements requirements
RUN pip install --upgrade pip setuptools wheel \
&& pip install -r requirements.txt
WORKDIR /app
COPY . .
# Set nvcc architecture
ENV GPU_TARGETS=${ROCM_DOCKER_ARCH}
# Enable ROCm
ENV LLAMA_HIPBLAS=1
ENV CC=/opt/rocm/llvm/bin/clang
ENV CXX=/opt/rocm/llvm/bin/clang++
RUN make
ENTRYPOINT [ "/app/main" ]