feat: fix CUDA images and update go-llama to use full GPU offloading (#618)

Signed-off-by: mudler <mudler@localai.io>
Co-authored-by: mudler <mudler@localai.io>
This commit is contained in:
Ettore Di Giacinto 2023-06-18 08:27:29 +02:00 committed by GitHub
parent 6c94f3cd67
commit d3d3187e51
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
5 changed files with 11 additions and 47 deletions

View file

@ -1,4 +1,4 @@
ARG GO_VERSION=1.20
ARG GO_VERSION=1.20-bullseye
FROM golang:$GO_VERSION as requirements
@ -9,7 +9,7 @@ ARG CUDA_MINOR_VERSION=7
ENV BUILD_TYPE=${BUILD_TYPE}
RUN apt-get update && \
apt-get install -y ca-certificates cmake curl
apt-get install -y ca-certificates cmake curl patch
# CuBLAS requirements
RUN if [ "${BUILD_TYPE}" = "cublas" ]; then \
@ -67,4 +67,4 @@ HEALTHCHECK --interval=1m --timeout=10m --retries=10 \
CMD curl -f $HEALTHCHECK_ENDPOINT || exit 1
EXPOSE 8080
ENTRYPOINT [ "/build/entrypoint.sh" ]
ENTRYPOINT [ "/build/entrypoint.sh" ]