matt/ollama37
1
0
Fork 0
mirror of https://github.com/dogkeeper886/ollama37.git synced 2025-12-19 20:27:01 +00:00
Code Issues Packages Projects Releases Wiki Activity

Support multiple variants for a given llm lib type

Browse Source 
In some cases we may want multiple variants for a given GPU type or CPU.
This adds logic to have an optional Variant which we can use to select
an optimal library, but also allows us to try multiple variants in case
some fail to load.

This can be useful for scenarios such as ROCm v5 vs v6 incompatibility
or potentially CPU features.
This commit is contained in:
Daniel Hiltgen
2024-01-05 12:13:08 -08:00
parent b24e8d17b2
commit 8da7bef05f
16 changed files with 428 additions and 212 deletions
Download Patch File Download Diff File Expand all files Collapse all files

52
Dockerfile.build
View File

@@ -1,7 +1,6 @@
ARG GOLANG_VERSION=1.21.3
ARG CMAKE_VERSION=3.22.1
ARG CUDA_VERSION=11.3.1
ARG ROCM_VERSION=5.7.1
FROM --platform=linux/amd64 nvidia/cuda:$CUDA_VERSION-devel-centos7 AS cuda-build-amd64
@@ -16,9 +15,11 @@ ADD https://github.com/Kitware/CMake/releases/download/v$CMAKE_VERSION/cmake-$CM
RUN tar -zx -C /usr --strip-components 1 </tmp/cmake-$CMAKE_VERSION.tar.gz
WORKDIR /go/src/github.com/jmorganca/ollama
COPY . .
COPY .git .git
COPY .gitmodules .gitmodules
COPY llm llm
WORKDIR llm/generate
WORKDIR /go/src/github.com/jmorganca/ollama/llm/generate
RUN sh gen_linux.sh
FROM --platform=linux/arm64 nvidia/cuda:$CUDA_VERSION-devel-rockylinux8 AS cuda-build-arm64
@@ -28,30 +29,46 @@ ARG CMAKE_VERSION
RUN dnf install -y git cmake
WORKDIR /go/src/github.com/jmorganca/ollama
COPY . .
COPY .git .git
COPY .gitmodules .gitmodules
COPY llm llm
WORKDIR llm/generate
WORKDIR /go/src/github.com/jmorganca/ollama/llm/generate
RUN sh gen_linux.sh
FROM --platform=linux/amd64 rocm/dev-centos-7:$ROCM_VERSION-complete AS rocm-build-amd64
FROM --platform=linux/amd64 rocm/dev-centos-7:5.7.1-complete AS rocm-5-build-amd64
ARG CMAKE_VERSION
RUN yum install -y https://repo.ius.io/ius-release-el7.rpm centos-release-scl \
&& yum update -y \
&& yum remove -y git \
&& yum install -y devtoolset-10-gcc devtoolset-10-gcc-c++ git236
ENV PATH /opt/rh/devtoolset-10/root/usr/bin:$PATH
ENV LIBRARY_PATH /opt/amdgpu/lib64
ADD https://github.com/Kitware/CMake/releases/download/v$CMAKE_VERSION/cmake-$CMAKE_VERSION-linux-x86_64.tar.gz /tmp/cmake-$CMAKE_VERSION.tar.gz
RUN tar -zx -C /usr --strip-components 1 </tmp/cmake-$CMAKE_VERSION.tar.gz
WORKDIR /go/src/github.com/jmorganca/ollama
COPY . .
COPY .git .git
COPY .gitmodules .gitmodules
COPY llm llm
WORKDIR /go/src/github.com/jmorganca/ollama/llm/generate
RUN OLLAMA_SKIP_CPU_GENERATE=1 sh gen_linux.sh
WORKDIR llm/generate
RUN sh gen_linux.sh
FROM --platform=linux/amd64 rocm/dev-centos-7:6.0-complete AS rocm-6-build-amd64
ARG CMAKE_VERSION
RUN yum install -y https://repo.ius.io/ius-release-el7.rpm centos-release-scl \
&& yum update -y \
&& yum remove -y git \
&& yum install -y devtoolset-10-gcc devtoolset-10-gcc-c++ git236
ENV PATH /opt/rh/devtoolset-10/root/usr/bin:$PATH
ENV LIBRARY_PATH /opt/amdgpu/lib64
ADD https://github.com/Kitware/CMake/releases/download/v$CMAKE_VERSION/cmake-$CMAKE_VERSION-linux-x86_64.tar.gz /tmp/cmake-$CMAKE_VERSION.tar.gz
RUN tar -zx -C /usr --strip-components 1 </tmp/cmake-$CMAKE_VERSION.tar.gz
WORKDIR /go/src/github.com/jmorganca/ollama
COPY .git .git
COPY .gitmodules .gitmodules
COPY llm llm
WORKDIR /go/src/github.com/jmorganca/ollama/llm/generate
RUN OLLAMA_SKIP_CPU_GENERATE=1 sh gen_linux.sh
FROM --platform=linux/amd64 centos:7 AS build-amd64
ENV CGO_ENABLED 1
@@ -71,9 +88,9 @@ ENV PATH /usr/local/go/bin:$PATH
WORKDIR /go/src/github.com/jmorganca/ollama
COPY . .
COPY --from=cuda-build-amd64 /go/src/github.com/jmorganca/ollama/llm/llama.cpp/build/linux/cpu/lib llm/llama.cpp/build/linux/cpu/lib
COPY --from=cuda-build-amd64 /go/src/github.com/jmorganca/ollama/llm/llama.cpp/build/linux/cuda/lib llm/llama.cpp/build/linux/cuda/lib
COPY --from=rocm-build-amd64 /go/src/github.com/jmorganca/ollama/llm/llama.cpp/build/linux/rocm/lib llm/llama.cpp/build/linux/rocm/lib
COPY --from=cuda-build-amd64 /go/src/github.com/jmorganca/ollama/llm/llama.cpp/build/linux/ llm/llama.cpp/build/linux/
COPY --from=rocm-5-build-amd64 /go/src/github.com/jmorganca/ollama/llm/llama.cpp/build/linux/ llm/llama.cpp/build/linux/
COPY --from=rocm-6-build-amd64 /go/src/github.com/jmorganca/ollama/llm/llama.cpp/build/linux/ llm/llama.cpp/build/linux/
RUN go build .
FROM --platform=linux/arm64 centos:7 AS build-arm64
@@ -94,8 +111,7 @@ ENV PATH /usr/local/go/bin:$PATH
WORKDIR /go/src/github.com/jmorganca/ollama
COPY . .
COPY --from=cuda-build-arm64 /go/src/github.com/jmorganca/ollama/llm/llama.cpp/build/linux/cpu/lib llm/llama.cpp/build/linux/cpu/lib
COPY --from=cuda-build-arm64 /go/src/github.com/jmorganca/ollama/llm/llama.cpp/build/linux/cuda/lib llm/llama.cpp/build/linux/cuda/lib
COPY --from=cuda-build-arm64 /go/src/github.com/jmorganca/ollama/llm/llama.cpp/build/linux/ llm/llama.cpp/build/linux/
RUN go build .
FROM build-$TARGETARCH