Compare commits

...

3 Commits

Author SHA1 Message Date
4729af5774 fix: update backup compose for OS switch compatibility
- restic/restic -> restic/rest-server (restic server subcommand removed)
- embergarage/restic-browser -> floibach/restic-browser (old image deprecated)
- mount Restic/ -> /repo for browser filesystem scanning
- add persistent browser-data volume for SQLite config DB
- remove obsolete version attribute
2026-07-05 21:34:54 -04:00
0ba694adfa fix: add user directive to hermes service 2026-07-05 21:34:54 -04:00
aec2ea8143 feat: add llama-cpp-hermes service with ROCm 6.1 + gfx906 support
- Add custom llama.cpp Dockerfile with ROCm 6.1 + gfx906 (MI50) build
- Add llama-cpp-hermes service serving Hermes 4.3 on dual MI50 GPUs
- Strip GPU devices/ROCm env from ollama service (CPU-only for embeddings)

Hermes 4.3 runs at ~19 t/s on dual MI50s with 160K context.
2026-07-05 21:34:54 -04:00
4 changed files with 77 additions and 123 deletions

View File

@@ -30,6 +30,7 @@ services:
ssh:
- default
container_name: hermes
user: "10000:10000"
entrypoint: ["/bin/bash", "-c",
"bash /opt/data/hermes-tools/install.sh && bash /usr/local/bin/run-multi-gateways.sh && exec /usr/bin/tini -g -- /opt/hermes/docker/entrypoint.sh \"$@\"",
"hermes-entrypoint"]
@@ -144,22 +145,7 @@ services:
- /mnt/HoardingCow_docker_data/Ollama/ollama:/root/.ollama
environment:
- OLLAMA_VULKAN=0
- HSA_OVERRIDE_GFX_VERSION=9.0.6
- HCC_AMDGPU_TARGET=gfx906
- HIP_VISIBLE_DEVICES=0,1
- ROCR_VISIBLE_DEVICES=0,1
- HSA_ENABLE_SDMA=0
- OLLAMA_HOST=0.0.0.0
- OLLAMA_DEBUG=1
- OLLAMA_FLASH_ATTENTION=1
- OLLAMA_NUM_PARALLEL=2
devices:
# Map the render nodes and KFD for ROCm to work inside the container
- /dev/kfd:/dev/kfd
- /dev/dri:/dev/dri
group_add:
- "303"
- "26"
# --- Honcho + OpenConcho combiné: API + Web UI nginx/FastAPI ---
honcho:
@@ -260,47 +246,40 @@ volumes:
external: true
name: honcho_data
# llama_cpp_devstral:
# image: ghcr.io/ggml-org/llama.cpp:server-rocm
# container_name: llama_cpp_devstral
# restart: unless-stopped
# networks:
# - ai_backend
# ports:
# - "8300:8080"
# ipc: host
# devices:
# - "/dev/kfd:/dev/kfd"
# - "/dev/dri:/dev/dri"
# group_add:
# - "303" # video
# - "26" # render
# environment:
# HSA_OVERRIDE_GFX_VERSION: 9.0.6
# HIP_VISIBLE_DEVICES: 0,1
# LLAMA_CACHE: /models
# volumes:
# - /mnt/HoardingCow_docker_data/Llama_cpp/models:/models
# - /mnt/HoardingCow_docker_data/Llama_cpp/devstral-agent.jinja:/template.jinja
# command: >
# -hf unsloth/Devstral-Small-2-24B-Instruct-2512-GGUF:Devstral-Small-2-24B-Instruct-2512-Q8_0.gguf
# -a devstral-2-small-llama_cpp
# --chat-template-file /template.jinja
# --host 0.0.0.0
# --port 8080
# --n-gpu-layers 99
# --ctx-size 163840
# --batch-size 4096
# --ubatch-size 4096
# --cache-type-k f16
# --cache-type-v f16
# --cache-reuse 256
# --flash-attn on
# --context-shift
# --split-mode layer
# --no-mmap
# --n-predict -1
# --parallel 2
llama-cpp-hermes:
image: llama-cpp:rocm-gfx906
container_name: llama-cpp-hermes
restart: unless-stopped
networks:
- ai_backend
ports:
- "127.0.0.1:8300:8080"
ipc: host
devices:
- /dev/kfd:/dev/kfd
- /dev/dri:/dev/dri
group_add:
- "303"
- "26"
environment:
- HSA_OVERRIDE_GFX_VERSION=9.0.6
- HSA_ENABLE_SDMA=0
- HIP_VISIBLE_DEVICES=0,1
- LLAMA_CACHE=/models
volumes:
- /mnt/HoardingCow_docker_data/Llama_cpp/models:/models
- /mnt/HoardingCow_docker_data/Ollama/ollama/models/blobs/sha256-17823599694fa3503ef54bf748d5078c6ce881f4d01616cafa255dc05d215a08:/model.gguf:ro
command: >
-m /model.gguf
--host 0.0.0.0
--port 8080
--gpu-layers 99
--ctx-size 163840
-ctk f16 -ctv f16
--flash-attn on
--split-mode layer
--no-mmap
--n-predict -1
# vllm:
# image: nalanzeyu/vllm-gfx906:v0.9.0-rocm6.3

30
ai/llama-cpp/Dockerfile Normal file
View File

@@ -0,0 +1,30 @@
# llama-cpp-rocm6/Dockerfile
# Custom llama.cpp server with ROCm 6.1 + gfx906 (MI50) support.
# Build: docker build -t llama-cpp:rocm-gfx906 .
FROM rocm/dev-ubuntu-22.04:6.1.2-complete AS builder
RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -y curl git build-essential pkg-config cmake make && rm -rf /var/lib/apt/lists/*
ARG LLAMACPP_VERSION=b9596
RUN git clone --depth 1 --branch ${LLAMACPP_VERSION} https://github.com/ggml-org/llama.cpp.git /build
WORKDIR /build
ENV HIP_PATH=/opt/rocm ROCM_PATH=/opt/rocm PATH=/opt/rocm/bin:/opt/rocm/llvm/bin:${PATH} CMAKE_PREFIX_PATH=/opt/rocm
RUN mkdir build && cd build && \
cmake .. -DGGML_HIP=ON -DCMAKE_BUILD_TYPE=Release \
-DAMDGPU_TARGETS="gfx906:xnack-" \
-DCMAKE_POSITION_INDEPENDENT_CODE=ON \
-DGGML_CUDA=OFF -DGGML_VULKAN=OFF -DGGML_METAL=OFF \
-DBUILD_SHARED_LIBS=OFF && \
cmake --build . --target llama-server -- -j $(nproc)
FROM ubuntu:24.04
RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -y \
ca-certificates curl libstdc++6 libgomp1 libopenblas0 \
libnuma1 libelf1 libdrm2 libdrm-amdgpu1 \
&& rm -rf /var/lib/apt/lists/*
COPY --from=builder /opt/rocm/lib/ /opt/rocm/lib/
COPY --from=builder /opt/rocm/share/ /opt/rocm/share/
COPY --from=builder /build/build/bin/llama-server /usr/local/bin/llama-server
RUN echo /opt/rocm/lib > /etc/ld.so.conf.d/rocm.conf && ldconfig
ENV HSA_OVERRIDE_GFX_VERSION=9.0.6 HCC_AMDGPU_TARGET=gfx906 HSA_ENABLE_SDMA=0
EXPOSE 8080
ENTRYPOINT ["/usr/local/bin/llama-server"]

1
backup/.env Normal file
View File

@@ -0,0 +1 @@
RESTIC_PASSWORD=CVvxpEKWSydmsZ5bW9lb_UZ2NR3KAKibv00tYbyZ9-E

View File

@@ -1,98 +1,42 @@
version: "3.8"
services:
# kopia:
# image: kopia/kopia:latest
# container_name: kopia
# restart: unless-stopped
# # We explicitly run as root (0:0) to solve the CHDIR issue,
# # OR we make sure the host folders match UID 1000.
# user: "0:0"
# command:
# - server
# - start
# - --address=0.0.0.0:51515
# - --server-username=${KOPIA_SERVER_USER}
# - --server-password=${KOPIA_SERVER_PASSWORD}
# - --config-file=/app/config/repository.config
# - --disable-csrf-token-checks
# - --insecure
# environment:
# - TZ=America/Montreal
# - KOPIA_PASSWORD=${KOPIA_PASSWORD}
# - USER=${KOPIA_USER}
# volumes:
# - /mnt/HoardingCow_docker_data/Kopia/config:/app/config
# - /mnt/HoardingCow_docker_data/Kopia/cache:/app/cache
# - /mnt/HoardingCow_docker_data/Kopia/repository:/repository
# # Required if you want to use the 'Mount' feature later
# - /tmp:/tmp:shared
# # Required for mounting backups as drives
# cap_add:
# - SYS_ADMIN
# devices:
# - /dev/fuse:/dev/fuse
# networks:
# - traefik-net
# labels:
# - "traefik.enable=true"
# # 1. HTTP to HTTPS Redirect
# - "traefik.http.routers.kopia-http.rule=Host(`backup.lazyworkhorse.net`)"
# - "traefik.http.routers.kopia-http.entrypoints=web"
# - "traefik.http.routers.kopia-http.middlewares=redirect-to-https@docker"
#
# # 2. HTTPS Configuration
# - "traefik.http.routers.kopia.rule=Host(`backup.lazyworkhorse.net`)"
# - "traefik.http.routers.kopia.entrypoints=websecure"
# - "traefik.http.routers.kopia.tls=true"
# - "traefik.http.routers.kopia.tls.certresolver=njalla"
#
# # 3. Backend Service Config
# - "traefik.http.services.kopia.loadbalancer.server.port=51515"
restic-server:
image: restic/restic:latest
image: restic/rest-server:latest
container_name: restic-server
restart: always
user: "0:0"
command: ["server", "--listen", ":8080", "--repo", "/data", "--tls-cert", "", "--tls-key", ""]
command: ["rest-server", "--listen", ":8080", "--path", "/data", "--no-auth"]
environment:
- TZ=America/Montreal
- RESTIC_PASSWORD=${RESTIC_PASSWORD}
volumes:
- /mnt/HoardingCow_docker_data/Restic/data:/data
# Mount paths to backup (adjust as needed)
- /mnt/HoardingCow_docker_data:/source:ro
networks:
- backup_net
labels:
- "traefik.enable=false" # Internal only, accessed by restic-browser
- "traefik.enable=false"
restic-browser:
image: embergarage/restic-browser:latest
image: floibach/restic-browser:latest
container_name: restic-browser
restart: always
restart: always
ports:
- "8081:8080"
environment:
- TZ=America/Montreal
- RESTIC_REPOSITORY=http://restic-server:8080
- RESTIC_PASSWORD=${RESTIC_PASSWORD}
volumes:
- /mnt/HoardingCow_docker_data/Restic:/repo:ro
- /mnt/HoardingCow_docker_data/Restic/browser-data:/data
networks:
- backup_net
labels:
- "traefik.enable=true"
# 1. HTTP to HTTPS Redirect
- "traefik.http.routers.restic-browser-http.rule=Host(`backup.lazyworkhorse.net`)"
- "traefik.http.routers.restic-browser-http.entrypoints=web"
- "traefik.http.routers.restic-browser-http.middlewares=redirect-to-https@docker"
# 2. HTTPS Configuration
- "traefik.http.routers.restic-browser.rule=Host(`backup.lazyworkhorse.net`)"
- "traefik.http.routers.restic-browser.entrypoints=websecure"
- "traefik.http.routers.restic-browser.tls=true"
- "traefik.http.routers.restic-browser.tls.certresolver=njalla"
# 3. Backend Service Config
- "traefik.http.services.restic-browser.loadbalancer.server.port=8000"
- "traefik.http.services.restic-browser.loadbalancer.server.port=8080"
networks:
backup_net: