Compare commits

..

21 Commits

Author SHA1 Message Date
46241d10c5 feat: update Hermes Dockerfile to build from forked source
- Switch Dockerfile to clone from gortium/hermes-agent (Gitea fork)
- Add SSH agent forwarding for private repo clone at build time
- Set CHROME_EXECUTABLE for Playwright Chromium
- Remove patch_tts_tool.py (Piper patch now in fork source)
- Enable Gitea Actions in versioncontrol compose
2026-05-10 17:55:17 -04:00
6b82a26c25 fix: add ldflags for version, remove privileged, enable flash attention 2026-05-10 10:07:25 -04:00
9cc7edfb39 fix: set CMAKE_INSTALL_PREFIX=/build/dist at configure time for CPU, match preset for HIP 2026-05-10 00:44:56 -04:00
f31ae59717 fix: copy /build/dist/lib/ollama/ (not /build/dist/lib/) to avoid extra nesting 2026-05-10 00:10:39 -04:00
bf2f17c5e2 fix: use cmake -B to override preset binaryDir, cmake --build/--install use explicit path 2026-05-09 23:52:46 -04:00
208bfd4612 fix: pre-set CMAKE_HIP_COMPILER="" for CPU build to prevent HIP detection 2026-05-09 23:50:26 -04:00
32df546550 fix: use ROCm 6 preset with HIP language detection for proper GPU kernel compilation
- Use --preset 'ROCm 6' for HIP build step (enables enable_language(HIP))
- Remove /opt/rocm from PATH for CPU build to prevent check_language(HIP)
- Add CMAKE_PREFIX_PATH=/opt/rocm so find_package(hip) finds hip-config.cmake
- cmake --install --component HIP now works correctly with OLLAMA_RUNNER_DIR=rocm
2026-05-09 23:49:08 -04:00
0f7b22c19b fix: add /usr/local/go/bin to ROCm PATH (was overridden)
ENV PATH for ROCm overwrote the previous PATH that included Go.
Without Go in PATH, 'go build' fails with 'go: not found'.
2026-05-09 23:15:26 -04:00
fc777e2de2 fix: target only gfx906 for HIP compilation
gfx940/gfx1010/gfx1030/gfx1100 cause C++ narrowing errors in ollama's
mma.cuh with hipcc. Since we only have MI50 (gfx906) cards, compile
for gfx906 only. Reduces build time and avoids upstream code bugs.
2026-05-09 23:07:39 -04:00
d52f18b0fa fix: remove gfx1200 target (not supported by ROCm 6.1 clang 17)
ROCm 6.1's AMD clang 17 doesn't recognize gfx1200 architecture
(introduced in ROCm 6.2+). Caused compilation failure on all .cu files.
2026-05-09 22:53:11 -04:00
0d87fb2556 fix: build CPU and HIP backends separately
CPU backends compiled with GCC (fixes AVX intrinsic errors from hipcc).
HIP backend compiled with hipcc (legacy mode skips enable_language(HIP)).
Go binary built with GCC for CGo linking.
This avoids both CMAKE_HIP_COMPILER rejection and CXX=hipcc CPU failures.
2026-05-09 22:51:13 -04:00
f6bc2b07a7 fix: remove nonexistent CC=clang for Go build step
ROCm 6.1 image doesn't have clang/clang++ in PATH (only amdclang++).
GCC is the default and works fine for CGo linking.
2026-05-09 22:41:18 -04:00
aa6bbe87bf fix: correct AMDGPU_TARGETS to include gfx940/gfx1010/gfx1200
Targets were corrupted during previous patch iterations, contained
gfx908/gfx90a from the CMake preset instead of gfx940/gfx1010/gfx1200.
2026-05-09 22:40:40 -04:00
0c612d9731 fix: remove unsupported AMDGPU_TARGETS (gfx1200) for ROCm 6.1
ROCm 6.1's AMD clang 17 doesn't support gfx1200 (RDNA4).
Use only targets supported by ROCm 6.1: gfx906, gfx908, gfx90a, gfx1030, gfx1100.
2026-05-09 22:30:21 -04:00
5b210fe624 fix: use ROCm amdclang++ as HIP compiler, keep GCC for CPU code
Setting CXX=hipcc caused compilation failures on CPU backends (AVX
intrinsics). Now using GCC for CPU, ROCm's amdclang++ for HIP only.
2026-05-09 22:29:10 -04:00
d8b77c97c3 fix: use CXX=hipcc legacy mode for HIP CMake build
CMake 3.31 refuses CMAKE_HIP_COMPILER=hipcc with 'not supported'.
Using CXX=hipcc triggers the legacy HIP detection path which works.
2026-05-09 22:20:44 -04:00
a3d0fa0072 fix: set CMAKE_HIP_COMPILER explicitly for ROCm 6.1 HIP detection 2026-05-09 22:19:50 -04:00
956d76f14d fix: add unzip dependency for ninja installation 2026-05-09 22:14:53 -04:00
c6d2f5918f fix: use ollama v0.23.2 native CMake build system for ROCm 6 + gfx906
The old Dockerfile used the deprecated llama.cpp/ subdirectory approach
which doesn't exist in ollama v0.23.2. Now using the official CMake
presets (ROCm 6 preset) with AMDGPU_TARGETS including gfx906:xnack-.
2026-05-09 22:13:47 -04:00
f023dc1ee4 fix: update ollama Dockerfile to v0.23.2 with proper ROCm 6.1 + gfx906 build
- Update OLLAMA_VERSION from v0.13.5 to v0.23.2
- Fix package: golang -> golang-go
- Add ENV HCC_AMDGPU_TARGET=gfx906 and HSA_ENABLE_SDMA=0
- Set proper ENTRYPOINT + CMD
2026-05-09 21:56:14 -04:00
d34a4d3647 refactor: move hermes files into ai/hermes/ subdirectory
- ai/Dockerfile -> ai/hermes/Dockerfile
- ai/fix-permissions.sh -> ai/hermes/fix-permissions.sh
- ai/patch_tts_tool.py -> ai/hermes/patch_tts_tool.py
- ai/compose.yml: update hermes build context to ./hermes
- ollama stays at ai/ollama/Dockerfile
2026-05-09 21:50:04 -04:00
11 changed files with 490 additions and 214 deletions

View File

@@ -2,7 +2,7 @@
COMPOSE_PATH=~/Projects/AltNet/docker-compose
# List of services (folder names)
SERVICES=monitoring hermes ollama cloudstorage crm_tp crm_cf mediacenter homeautomation network backup homepage passwordmanager
SERVICES=monitoring ai cloudstorage crm_tp crm_cf mediacenter homeautomation network backup homepage passwordmanager
# Bring up all services
all_up:

305
ai/compose.yml Normal file
View File

@@ -0,0 +1,305 @@
version: "3.8"
services:
# webui:
# image: ghcr.io/open-webui/open-webui:main
# volumes:
# - /mnt/HoardingCow_docker_data/Ollama/open-webui:/app/backend/data
# restart: always
# environment:
# - OLLAMA_API_BASE_URL=http://ollama:11434/api
# networks:
# - ai_net
# - ai_backend
# labels:
# - "traefik.enable=true"
# # Router for HTTP + redirection to HTTPS
# - "traefik.http.routers.webui-http.rule=Host(`ai.lazyworkhorse.net`)"
# - "traefik.http.routers.webui-http.entrypoints=web"
# - "traefik.http.routers.webui-http.middlewares=redirect-to-https"
# # Router for HTTPS with TLS
# - "traefik.http.routers.webui-https.rule=Host(`ai.lazyworkhorse.net`)"
# - "traefik.http.routers.webui-https.entrypoints=websecure"
# - "traefik.http.routers.webui-https.tls=true"
# - "traefik.http.routers.webui-https.tls.certresolver=njalla"
hermes:
build:
context: ./hermes
ssh:
- default
container_name: hermes
restart: always
# Gateway run enables the internal API server on port 8642
command: gateway run
environment:
- OLLAMA_HOST=http://ollama:11434
- API_SERVER_ENABLED=true
- API_SERVER_PORT=8642
- API_SERVER_HOST=0.0.0.0
- API_SERVER_KEY=hermes_local_key
- GATEWAY_ALLOW_ALL_USERS=true
- OPENROUTER_API_KEY=${OPENROUTER_API_KEY}
# ROCm for GPU-accelerated faster-whisper STT
- 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
- TZ=America/Montreal
volumes:
- /mnt/HoardingCow_docker_data/Hermes/data:/opt/data
devices:
- /dev/kfd:/dev/kfd
- /dev/dri:/dev/dri
group_add:
- "303"
- "26"
networks:
- ai_backend
ollama:
build:
context: ./ollama
dockerfile: Dockerfile
image: ollama/ollama:rocm-gfx906
container_name: ollama
tty: true
restart: always
ports:
- "127.0.0.1:11434:11434"
networks:
- ai_backend
volumes:
- /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"
networks:
ai_net:
external: true
name: ai_net
ai_backend:
driver: bridge
name: ai_backend
# 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
# vllm:
# image: nalanzeyu/vllm-gfx906:v0.9.0-rocm6.3
# container_name: vllm
# # Required for multi-GPU communication (NCCL)
# ipc: host
# init: true
# shm_size: '2g'
# networks:
# - ai_backend
# ports:
# - "8300:8000"
# 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
# NCCL_P2P_DISABLE: 1
# VLLM_WORKER_MULTIPROC_METHOD: spawn
# VLLM_USE_TRITON_FLASH_ATTN: 0
# VLLM_USE_ROCM_CUSTOM_PAGED_ATTN: 0
# VLLM_ATTENTION_BACKEND: ROPE_NAIVE
# VLLM_SKIP_WARMUP: 1
# VLLM_USE_V1: 0
# HF_TOKEN: ${HF_TOKEN}
# command: >
# vllm serve "mistralai/Devstral-Small-2-24B-Instruct-2512"
# --tensor-parallel-size 2
# --max-model-len 8192
# --gpu-memory-utilization 0.90
# --tokenizer_mode mistral
# --config_format auto
# --load-format auto
# --enforce-eager
# --disable-custom-all-reduce
# --trust-remote-code
# --task generate
# --block-size 16
# volumes:
# - /mnt/HoardingCow_docker_data/vllm/models:/root/.cache/huggingface
# restart: unless-stopped
# n8n:
# image: n8nio/n8n:latest
# container_name: n8n
# restart: unless-stopped
# networks:
# - ai_net
# environment:
# - N8N_HOST=n8n.lazyworkhorse.net
# - N8N_PORT=5678
# - N8N_PROTOCOL=https
# - NODE_ENV=production
# - N8N_ENCRYPTION_KEY=${N8N_ENCRYPTION_KEY}
# - WEBHOOK_URL=https://n8n.lazyworkhorse.net/
# - GENERIC_TIMEZONE=America/New_York # Adjust to your timezone
# - N8N_BLOCK_EXTERNAL_STORAGE_ACCESS=false
# - N8N_NODES_PYTHON_CAN_IMPORT_MODULES=true
# - N8N_NATIVE_PYTHON_RUNNER=true
# - N8N_PYTHON_ALLOW_STDLIB=uuid,re,os,json
# - N8N_PYTHON_ALLOW_EXTERNAL=requests,pandas
# - NODE_FUNCTION_ALLOW_EXTERNAL=uuid,requests
# volumes:
# - /mnt/HoardingCow_docker_data/n8n:/home/node/.n8n
# labels:
# - "traefik.enable=true"
# # Router for HTTP + redirection to HTTPS
# - "traefik.http.routers.n8n-http.rule=Host(`n8n.lazyworkhorse.net`)"
# - "traefik.http.routers.n8n-http.entrypoints=web"
# - "traefik.http.routers.n8n-http.middlewares=redirect-to-https"
# # Router for HTTPS with TLS
# - "traefik.http.routers.n8n-https.rule=Host(`n8n.lazyworkhorse.net`)"
# - "traefik.http.routers.n8n-https.entrypoints=websecure"
# - "traefik.http.routers.n8n-https.tls=true"
# - "traefik.http.routers.n8n-https.tls.certresolver=njalla"
# # Service Loadbalancer (n8n default port)
# - "traefik.http.services.n8n.loadbalancer.server.port=5678"
# openclaw:
# image: coollabsio/openclaw:latest
# container_name: openclaw
# restart: unless-stopped
# expose:
# - "8080" # WebUI
# - "18789" # Gateway/WebSocket
# - "8788" # Nextcloud Webhook
# networks:
# - ai_net
# - ai_backend
# volumes:
# - /mnt/HoardingCow_docker_data/openclaw/data:/data
# - /home/gortium/infra:/data/workspace/infra
# environment:
# - TZ=America/Toronto
# - OPENCLAW_GATEWAY_TOKEN=${OPENCLAW_GATEWAY_TOKEN}
# - OPENROUTER_API_KEY=${OPENROUTER_API_KEY}
# # Point to the sidecar browser
# - BROWSER_CDP_URL=http://openclaw-browser:9222
# - BROWSER_EVALUATE_ENABLED=true
# - OPENCLAW_GATEWAY_HOST=0.0.0.0
# - OPENCLAW_ALLOWED_ORIGINS=https://claw.lazyworkhorse.net
# labels:
# - "traefik.enable=true"
# - "traefik.http.routers.openclaw-http.rule=Host(`claw.lazyworkhorse.net`)"
# - "traefik.http.routers.openclaw-http.entrypoints=web"
# - "traefik.http.routers.openclaw-http.middlewares=redirect-to-https"
# - "traefik.http.routers.openclaw-https.rule=Host(`claw.lazyworkhorse.net`)"
# - "traefik.http.routers.openclaw-https.priority=50"
# - "traefik.http.routers.openclaw-https.entrypoints=websecure"
# - "traefik.http.routers.openclaw-https.tls=true"
# - "traefik.http.routers.openclaw-https.tls.certresolver=njalla"
# - "traefik.http.services.openclaw.loadbalancer.server.port=8080"
# depends_on:
# - openclaw-browser
# openclaw-browser:
# image: ghcr.io/browserless/chromium:latest
# restart: always
# expose:
# - "3000"
# environment:
# - MAX_CONCURRENT_SESSIONS=10
# - CONNECTION_TIMEOUT=300000
# - PREBOOT_CHROME=true
# - DEMO_MODE=false
# networks:
# ai_backend:
# aliases:
# - browser
# openclaw-ssh:
# image: linuxserver/openssh-server:latest
# container_name: openclaw-ssh
# environment:
# - PUID=1000
# - PGID=1000
# - PUBLIC_KEY_FILE=/config/ssh/authorized_keys
# - SUDO_ACCESS=false
# - PASSWORD_ACCESS=false
# volumes:
# - /mnt/HoardingCow_docker_data/openclaw/ssh-config:/config
# - /home/gortium/infra:/data/workspace/infra:ro
# restart: unless-stopped
# networks:
# - ai_backend
# labels:
# - "traefik.enable=true"
# - "traefik.tcp.routers.openclaw-ssh.rule=HostSNI(*)"
# - "traefik.tcp.routers.openclaw-ssh.entrypoints=sshnode"
# - "traefik.tcp.routers.openclaw-ssh.tls.passthrough=false"
# - "traefik.tcp.services.openclaw-ssh.loadbalancer.server.port=2222"

74
ai/hermes/Dockerfile Normal file
View File

@@ -0,0 +1,74 @@
# syntax=docker/dockerfile:1
# Hermes Agent -- custom fork build
# Builds on top of official image + overlays our forked source from Gitea.
# Requires Docker BuildKit. Pass SSH agent for git clone:
# docker compose build hermes
# Or manually:
# DOCKER_BUILDKIT=1 docker build --ssh default -t hermes-agent:custom .
# ---------- Base: official Hermes image (system deps, npm, uv, Playwright) ----------
FROM nousresearch/hermes-agent:latest
# ---------- Overlay our forked source ----------
# Uses SSH agent forwarding from the build host (no key baked into image).
# --exclude node_modules/.venv keeps the base image's pre-built layers intact.
# Only the Python source, web UI source, and config change.
RUN --mount=type=ssh \
mkdir -p /root/.ssh && \
ssh-keyscan -p 2222 code.lazyworkhorse.net >> /root/.ssh/known_hosts 2>/dev/null && \
cd /tmp && \
GIT_SSH_COMMAND='ssh -p 2222 -o StrictHostKeyChecking=no' \
git clone --depth 1 --branch main \
git@code.lazyworkhorse.net:gortium/hermes-agent.git fork && \
rsync -a --delete fork/ /opt/hermes/ \
--exclude node_modules \
--exclude .venv \
--exclude .git && \
rm -rf /tmp/fork /root/.ssh/
# ---------- Rebuild web UI ----------
# Source files changed; node_modules (from base image) reused.
RUN cd /opt/hermes && npm run build
# ---------- Reinstall Python package (editable) ----------
# Picks up source changes from our fork.
RUN . /opt/hermes/.venv/bin/activate && \
uv pip install --no-cache-dir --no-deps -e /opt/hermes
# ---------- Extra system deps ----------
USER root
RUN apt-get update && \
apt-get install -y --no-install-recommends \
libportaudio2 ca-certificates poppler-utils imagemagick \
texlive-latex-base texlive-latex-extra texlive-fonts-recommended \
texlive-xetex texlive-science \
qemu-user-static binfmt-support emacs-nox && \
rm -rf /var/lib/apt/lists/*
# ---------- UV ----------
COPY --chmod=0755 --from=ghcr.io/astral-sh/uv:latest /uv /usr/local/bin/
# ---------- Piper TTS ----------
RUN . /opt/hermes/.venv/bin/activate && \
uv pip install --no-cache-dir piper-tts sounddevice numpy && \
mkdir -p /opt/hermes/.venv/share/piper/voices
RUN /opt/hermes/.venv/bin/python3 /dev/stdin << 'PYEOF'
import urllib.request
base = '/opt/hermes/.venv/share/piper/voices'
url = 'https://huggingface.co/rhasspy/piper-voices/resolve/main/en/en_US/ryan/high/en_US-ryan-high.onnx'
urllib.request.urlretrieve(url, base + '/en_US-ryan-high.onnx')
urllib.request.urlretrieve(url + '.json', base + '/en_US-ryan-high.onnx.json')
PYEOF
# ---------- Runtime ----------
USER hermes
ENV HERMES_HOME=/opt/data
ENV PATH="/opt/data/.local/bin:${PATH}"
# Point browser tool to Playwright's Chromium (already in base image)
ENV CHROME_EXECUTABLE=/opt/hermes/.playwright/chromium/chrome-linux/chrome
VOLUME [ "/opt/data" ]
COPY --chmod=0755 fix-permissions.sh /opt/hermes/fix-permissions.sh
ENTRYPOINT [ "/usr/bin/tini", "-g", "--", "/opt/hermes/fix-permissions.sh" ]

106
ai/ollama/Dockerfile Normal file
View File

@@ -0,0 +1,106 @@
# ollama-gfx906/Dockerfile
#
# Custom ollama image with ROCm 6.1 + gfx906 (MI50) support.
# The official ollama/rocm image ships ROCm 7.2 which dropped gfx906.
# This uses v0.23.2's native CMake build system with AMDGPU_TARGETS including gfx906.
#
# Build: docker build -t ollama/ollama:rocm-gfx906 ai/ollama
FROM rocm/dev-ubuntu-22.04:6.1.2-complete AS builder
# Build dependencies (CMake, Ninja, Go)
ARG CMAKEVERSION=3.31.2
ARG NINJAVERSION=1.12.1
ARG GOLANG_VERSION=1.22.0
RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -y \
curl git ccache build-essential pkg-config unzip \
&& rm -rf /var/lib/apt/lists/*
# Install CMake from official binaries
RUN curl -fsSL https://github.com/Kitware/CMake/releases/download/v${CMAKEVERSION}/cmake-${CMAKEVERSION}-linux-x86_64.tar.gz \
| tar xz -C /usr/local --strip-components 1
# Install Ninja
RUN curl -fsSL -o /tmp/ninja.zip \
https://github.com/ninja-build/ninja/releases/download/v${NINJAVERSION}/ninja-linux.zip \
&& unzip /tmp/ninja.zip -d /usr/local/bin && rm /tmp/ninja.zip
# Install Go
RUN curl -fsSL https://go.dev/dl/go${GOLANG_VERSION}.linux-amd64.tar.gz \
| tar xz -C /usr/local
ENV PATH=/usr/local/go/bin:$PATH
ARG OLLAMA_VERSION=v0.23.2
RUN git clone --depth 1 --branch ${OLLAMA_VERSION} https://github.com/ollama/ollama.git /build
WORKDIR /build
# ROCm paths
ENV HIP_PATH=/opt/rocm
ENV ROCM_PATH=/opt/rocm
ENV CMAKE_GENERATOR=Ninja
ENV LDFLAGS=-s
# Step 1: Build CPU backends with GCC (no ROCm preset)
# Pre-set CMAKE_HIP_COMPILER="" to prevent check_language(HIP) from
# finding a HIP compiler (it searches /opt/rocm even without PATH).
# Remove /opt/rocm from PATH to prevent find_program from finding hipcc.
RUN mkdir -p build-cpu && \
PATH=/usr/local/go/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin \
cmake -B build-cpu -DCMAKE_BUILD_TYPE=Release \
-DCMAKE_HIP_COMPILER="" \
-DCMAKE_INSTALL_PREFIX=/build/dist && \
cmake --build build-cpu --target ggml-cpu -- -l $(nproc) && \
cmake --install build-cpu --component CPU --strip && \
echo "=== CPU install ===" && \
(find /build/dist/lib/ollama -type f -o -type l 2>&1 | head -20 || echo "empty")
# Step 2: Build HIP backend with ROCm preset + gfx906 target only
# The ROCm 6 preset enables HIP language detection (enable_language(HIP))
# which ensures GPU kernels are properly compiled for gfx906.
# OLLAMA_RUNNER_DIR=rocm from the preset, so HIP goes to lib/ollama/rocm/
# Need CMAKE_PREFIX_PATH so find_package(hip) finds hip-config.cmake
# at /opt/rocm/lib/cmake/hip/hip-config.cmake.
RUN mkdir -p build-hip && \
cmake -B build-hip \
--preset 'ROCm 6' \
-DAMDGPU_TARGETS="gfx906:xnack-" \
-DCMAKE_BUILD_TYPE=Release \
-DCMAKE_PREFIX_PATH="/opt/rocm" && \
cmake --build build-hip --target ggml-hip -- -l $(nproc) && \
cmake --install build-hip --component HIP --strip && \
echo "=== HIP install ===" && \
find /build/dist/lib/ollama -type f -o -type l | head -20
# Step 3: Build Go binary (GCC for CGo linking)
ENV CGO_ENABLED=1
RUN go build -trimpath -ldflags="-X=github.com/ollama/ollama/version.Version=${OLLAMA_VERSION}" -o /build/dist/ollama .
# ---------- Runtime image ----------
FROM ubuntu:24.04
RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -y \
ca-certificates curl libstdc++6 libgomp1 libvulkan1 libopenblas0 \
&& rm -rf /var/lib/apt/lists/*
# Copy ROCm 6.1 runtime libraries
# These are needed at runtime by ggml-hip via LD_LIBRARY_PATH
COPY --from=builder /opt/rocm/lib/ /opt/rocm/lib/
COPY --from=builder /opt/rocm/share/ /opt/rocm/share/
# Copy ollama binary + all backends (CPU + HIP)
# CPU install: /build/dist/lib/ollama/libggml-*.so
# HIP install: /build/dist/lib/ollama/rocm/libggml-hip.so
COPY --from=builder /build/dist/ollama /usr/bin/ollama
COPY --from=builder /build/dist/lib/ollama/ /usr/lib/ollama/
RUN ldconfig
ENV LD_LIBRARY_PATH=/opt/rocm/lib:/usr/lib/ollama/rocm:/usr/lib/ollama
ENV HSA_OVERRIDE_GFX_VERSION=9.0.6
ENV HCC_AMDGPU_TARGET=gfx906
ENV HSA_ENABLE_SDMA=0
EXPOSE 11434
ENTRYPOINT ["/bin/ollama"]
CMD ["serve"]

View File

@@ -1,73 +0,0 @@
# 1. On récupère la version la plus récente d'UV
FROM ghcr.io/astral-sh/uv:latest AS uv_source
# 2. Image officielle Hermes Agent de NousResearch
# Contient déjà: Python, Node.js, npm, Playwright/Chromium, venv, tts_tool.py, etc.
FROM nousresearch/hermes-agent:latest
# ---------- System dependencies ----------
# The official hermes-agent image already has: git, curl, ffmpeg, python3,
# gcc, build-essential, openssh-client, procps, tini, ripgrep, docker-cli,
# libportaudio2, ca-certificates, etc.
#
# These extras we need to add back:
# - poppler-utils, imagemagick (PDF/image processing)
# - texlive-* (LaTeX typesetting for reports)
# - qemu-user-static, binfmt-support (QEMU cross-compilation)
# - emacs-nox (text editing in container)
USER root
RUN apt-get update && \
apt-get install -y --no-install-recommends \
libportaudio2 \
ca-certificates \
poppler-utils \
imagemagick \
texlive-latex-base \
texlive-latex-extra \
texlive-fonts-recommended \
texlive-xetex \
texlive-science \
qemu-user-static \
binfmt-support \
emacs-nox && \
rm -rf /var/lib/apt/lists/*
# ---------- UV (hyperfast pip alternative) ----------
COPY --chmod=0755 --from=uv_source /uv /usr/local/bin/
WORKDIR /opt/hermes
# ---------- Piper TTS dans le venv existant ----------
# Le venv de l'image de base est root-owned, on doit installer en root aussi
RUN . /opt/hermes/.venv/bin/activate && \
uv pip install --no-cache-dir piper-tts sounddevice numpy
# ---------- Télécharger la voix Piper Ryan (high quality) ----------
RUN mkdir -p /opt/hermes/.venv/share/piper/voices && \
/opt/hermes/.venv/bin/python3 /dev/stdin << 'PYEOF'
import urllib.request
base = '/opt/hermes/.venv/share/piper/voices'
url = 'https://huggingface.co/rhasspy/piper-voices/resolve/main/en/en_US/ryan/high/en_US-ryan-high.onnx'
urllib.request.urlretrieve(url, base + '/en_US-ryan-high.onnx')
urllib.request.urlretrieve(url + '.json', base + '/en_US-ryan-high.onnx.json')
PYEOF
# ---------- Patch tts_tool.py: remplacer Edge TTS par Piper ----------
# Edge TTS appelle les serveurs Microsoft — on ne veut jamais ça.
# Piper roule localement sur CPU, aucun cloud, aucune donnée qui sort.
COPY patch_tts_tool.py /tmp/patch_tts_tool.py
RUN /opt/hermes/.venv/bin/python3 /tmp/patch_tts_tool.py && rm /tmp/patch_tts_tool.py
# ---------- Runtime ----------
# Retour à l'utilisateur non-root pour la sécurité
USER hermes
ENV HERMES_HOME=/opt/data
ENV PATH="/opt/data/.local/bin:${PATH}"
VOLUME [ "/opt/data" ]
# Script de réparation des permissions + patch TTS au démarrage
COPY --chmod=0755 fix-permissions.sh /opt/hermes/fix-permissions.sh
ENTRYPOINT [ "/usr/bin/tini", "-g", "--", "/opt/hermes/fix-permissions.sh" ]

View File

@@ -1,38 +0,0 @@
version: "3.8"
services:
hermes:
build: ./
container_name: hermes
restart: always
# Gateway run enables the internal API server on port 8642
command: gateway run
environment:
- OLLAMA_HOST=http://ollama:11434
- API_SERVER_ENABLED=true
- API_SERVER_PORT=8642
- API_SERVER_HOST=0.0.0.0
- API_SERVER_KEY=hermes_local_key
- GATEWAY_ALLOW_ALL_USERS=true
- OPENROUTER_API_KEY=${OPEN...KEY}
# ROCm for GPU-accelerated faster-whisper STT
- 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
volumes:
- /mnt/HoardingCow_docker_data/Hermes/data:/opt/data
devices:
- /dev/kfd:/dev/kfd
- /dev/dri:/dev/dri
group_add:
- "303"
- "26"
networks:
- ai_backend
networks:
ai_backend:
external: true
name: ai_backend

View File

@@ -1,61 +0,0 @@
# ollama/Dockerfile
#
# Custom ollama image with ROCm + gfx906 (MI50) support.
# The default ollama/rocm image ships ROCm 7.2 which drops gfx906 support.
# This builds ollama and its llama.cpp runner from source, targeting gfx906.
#
# Build:
# docker build -t ollama/ollama:rocm-gfx906 ./ollama
FROM rocm/dev-ubuntu-22.04:6.1.2-complete AS builder
RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -y \
git golang-go cmake build-essential pkg-config \
&& rm -rf /var/lib/apt/lists/*
ARG OLLAMA_VERSION=v0.23.2
RUN git clone --depth 1 --branch ${OLLAMA_VERSION} https://github.com/ollama/ollama.git /build
WORKDIR /build
ENV HIP_PATH=/opt/rocm
ENV ROCM_PATH=/opt/rocm
ENV PATH=/opt/rocm/bin:/opt/rocm/hip/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin
RUN cd llama.cpp && \
mkdir -p build && cd build && \
cmake .. \
-DLLAMA_HIPBLAS=ON \
-DCMAKE_C_COMPILER=clang \
-DCMAKE_CXX_COMPILER=clang++ \
-DAMDGPU_TARGETS=gfx906 \
-DCMAKE_BUILD_TYPE=Release \
-DLLAMA_NATIVE=OFF \
-DLLAMA_BUILD_TESTS=OFF \
-DLLAMA_BUILD_EXAMPLES=OFF \
-DLLAMA_BUILD_SERVER=OFF && \
cmake --build . --config Release -j$(nproc) && \
cmake --install . --prefix /build/dist
ENV CGO_ENABLED=0
RUN go build -trimpath -o dist/ollama .
FROM ubuntu:22.04
RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -y \
ca-certificates curl libstdc++6 libgomp1 \
&& 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/dist/ollama /usr/bin/ollama
COPY --from=builder /build/dist/lib/ /usr/lib/ollama/
RUN ldconfig
ENV LD_LIBRARY_PATH=/opt/rocm/lib:/usr/lib/ollama
ENV HSA_OVERRIDE_GFX_VERSION=9.0.6
ENV HCC_AMDGPU_TARGET=gfx906
ENV HSA_ENABLE_SDMA=0
EXPOSE 11434
ENTRYPOINT ["/bin/ollama"]
CMD ["serve"]

View File

@@ -1,41 +0,0 @@
version: "3.8"
services:
ollama:
build:
context: ./
dockerfile: Dockerfile
image: ollama/ollama:rocm-gfx906
container_name: ollama
privileged: true
tty: true
restart: always
ports:
- "127.0.0.1:11434:11434"
networks:
- ai_backend
volumes:
- /mnt/HoardingCow_docker_data/Ollama/ollama:/root/.ollama
environment:
- OLLAMA_VULKAN=0
- OLLAMA_HOST=0.0.0.0
- OLLAMA_DEBUG=1
- OLLAMA_FLASH_ATTENTION=0
- OLLAMA_NUM_PARALLEL=2
# ROCm / gfx906 configuration
- 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
devices:
- /dev/kfd:/dev/kfd
- /dev/dri:/dev/dri
group_add:
- "303"
- "26"
networks:
ai_backend:
external: true
name: ai_backend

View File

@@ -9,6 +9,10 @@ services:
- GITEA__server__ROOT_URL=https://code.lazyworkhorse.net
- SSH_PORT=2222
- SSH_LISTEN_PORT=2222
# Enable Gitea Actions (act_runner required on host)
- GITEA__actions__ENABLED=true
# Don't fetch actions from GitHub (offline mode + local only)
- GITEA__actions__DEFAULT_ACTIONS_URL=off
volumes:
- /mnt/HoardingCow_docker_data/Gitea:/data
networks: