Koko210/miku-discord
1
0
Fork 0
Code Issues 34 Pull Requests Actions Packages Projects Releases Wiki Activity

llama-swap-rocm now uses official image and adjusted accordingly

Browse Source
This commit is contained in:
koko210Serve
2026-02-07 23:43:01 +02:00
parent 11b90ebb46
commit b9d1f67d70
1 changed files with 27 additions and 27 deletions
Download Patch File Download Diff File Expand all files Collapse all files

14
Dockerfile.llamaswap-rocm
View File

@@ -39,23 +39,22 @@ WORKDIR /app
# Copy llama-swap binary from builder # Copy llama-swap binary from builder
COPY --from=swap-builder /build/llama-swap-binary /app/llama-swap COPY --from=swap-builder /build/llama-swap-binary /app/llama-swap
# Make binary executable # Make binaries executable
RUN chmod +x /app/llama-swap RUN chmod +x /app/llama-swap
# Create non-root user and add to GPU access groups # Add existing ubuntu user (UID 1000) to GPU access groups (using host GIDs)
# The official llama.cpp image already has llama-server installed
# GID 187 = render group on host, GID 989 = video/kfd group on host # GID 187 = render group on host, GID 989 = video/kfd group on host
RUN groupadd -g 187 hostrender && \ RUN groupadd -g 187 hostrender && \
groupadd -g 989 hostvideo && \ groupadd -g 989 hostvideo && \
useradd -m -u 1000 -G hostrender,hostvideo llamaswap && \ usermod -aG hostrender,hostvideo ubuntu && \
chown -R llamaswap:llamaswap /app chown -R ubuntu:ubuntu /app
# Set environment for ROCm (RX 6800 is gfx1030) # Set environment for ROCm (RX 6800 is gfx1030)
ENV HSA_OVERRIDE_GFX_VERSION=10.3.0 ENV HSA_OVERRIDE_GFX_VERSION=10.3.0
ENV ROCM_PATH=/opt/rocm ENV ROCM_PATH=/opt/rocm
ENV HIP_VISIBLE_DEVICES=0 ENV HIP_VISIBLE_DEVICES=0
USER llamaswap USER ubuntu
# Expose port # Expose port
EXPOSE 8080 EXPOSE 8080
@@ -64,5 +63,6 @@ EXPOSE 8080
HEALTHCHECK --interval=30s --timeout=10s --start-period=30s --retries=3 \ HEALTHCHECK --interval=30s --timeout=10s --start-period=30s --retries=3 \
CMD curl -f http://localhost:8080/health || exit 1 CMD curl -f http://localhost:8080/health || exit 1
# Run llama-swap # Override the base image's ENTRYPOINT and run llama-swap
ENTRYPOINT []
CMD ["/app/llama-swap", "-config", "/app/config.yaml", "-listen", "0.0.0.0:8080"] CMD ["/app/llama-swap", "-config", "/app/config.yaml", "-listen", "0.0.0.0:8080"]