EchOnyx/docker-compose.amd.yml at main · VACInc/EchOnyx · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
# AMD GPU override (Strix Halo / ROCm) - use with:
# docker compose -f docker-compose.yml -f docker-compose.amd.yml up

x-rocm-openai-runtime: &rocm-openai-runtime
  build:
    context: .
    dockerfile: docker/rocm-openai-runtime/Dockerfile
    args:
      ROCM_DEV_IMAGE: ${ROCM_DEV_IMAGE:-rocm/dev-ubuntu-24.04:7.2-complete}
      AMD_LLAMA_PACKAGE_URL: ${AMD_LLAMA_PACKAGE_URL:-https://repo.radeon.com/rocm/llama.cpp/linux/rocm-rel-7.2/llama-b7782-ubuntu-24.04-rocm-7.2.0-gfx110X-gfx115X-gfx120X-x64.zip}
      INSTALL_VLLM: ${INSTALL_VLLM:-0}
      VLLM_INSTALL_METHOD: ${VLLM_INSTALL_METHOD:-wheel}
      VLLM_WHEEL_EXTRA_INDEX_URL: ${VLLM_WHEEL_EXTRA_INDEX_URL:-https://wheels.vllm.ai/rocm/}
      ROCM_WHL_URL: ${ROCM_WHL_URL:-https://repo.radeon.com/rocm/manylinux/rocm-rel-7.2/}
      ROCM_TORCH_VERSION: ${ROCM_TORCH_VERSION:-2.10.0+rocm7.2.0.lw.gitb6ee5fde}
      ROCM_TORCHAUDIO_VERSION: ${ROCM_TORCHAUDIO_VERSION:-2.10.0+rocm7.2.0.git5047768f}
      ROCM_TORCHVISION_VERSION: ${ROCM_TORCHVISION_VERSION:-0.25.0+rocm7.2.0.git82df5f59}
      ROCM_NIGHTLY: ${ROCM_NIGHTLY:-0}
  restart: unless-stopped
  devices:
    - /dev/kfd:/dev/kfd
    - /dev/dri:/dev/dri
  group_add:
    - "44"   # video group
    - "993"  # render group
  environment:
    - HSA_OVERRIDE_GFX_VERSION=${HSA_OVERRIDE_GFX_VERSION:-11.5.1}
    - HF_HOME=/cache/huggingface
    - HUGGINGFACE_HUB_CACHE=/cache/huggingface/hub
  volumes:
    - model_cache:/models:ro
    - vllm_cache:/cache
  healthcheck:
    test: ["CMD", "curl", "-f", "http://localhost:8080/health"]
    interval: 30s
    timeout: 10s
    retries: 20
    start_period: 90s

services:
  vision-server:
    <<: *rocm-openai-runtime
    container_name: vs-vision
    environment:
      - MODEL_RUNTIME=${ROCM_LLM_RUNTIME:-llama_server}
      - IDLE_TIMEOUT_SECONDS=${ROCM_LLM_IDLE_TIMEOUT_S:-120}
      - MODEL_PATH=${VISION_MODEL_PATH:-/models/Qwen3VL-32B-Instruct-Q4_K_M.gguf}
      - MODEL_NAME=${VISION_ENDPOINT_MODEL:-Qwen3VL-32B-Instruct-Q4_K_M.gguf}
      - MODEL_MMPROJ=${VISION_MMPROJ_PATH:-/models/mmproj-Qwen3VL-32B-Instruct-F16.gguf}
      - VLLM_MODEL_ID=${VISION_VLLM_MODEL_ID:-Qwen/Qwen3-VL-30B-A3B-Instruct-FP8}
      - MODEL_CONTEXT_SIZE=${VISION_CONTEXT_SIZE:-8192}
      - MODEL_GPU_LAYERS=${VISION_GPU_LAYERS:-999}
      - LLAMA_SERVER_EXTRA_ARGS=${VISION_LLAMA_SERVER_EXTRA_ARGS:-}
      - VLLM_EXTRA_ARGS=${VISION_VLLM_EXTRA_ARGS:-}

  summarization-server:
    <<: *rocm-openai-runtime
    container_name: vs-summary
    environment:
      - MODEL_RUNTIME=${ROCM_LLM_RUNTIME:-llama_server}
      - IDLE_TIMEOUT_SECONDS=${ROCM_LLM_IDLE_TIMEOUT_S:-120}
      - MODEL_PATH=${SUMMARIZATION_MODEL_PATH:-/models/qwen3-30b-a3b-q4_k_m.gguf}
      - MODEL_NAME=${SUMMARIZATION_ENDPOINT_MODEL:-qwen3-30b-a3b-q4_k_m.gguf}
      - VLLM_MODEL_ID=${SUMMARIZATION_VLLM_MODEL_ID:-Qwen/Qwen3-30B-A3B-Instruct-2507-FP8}
      - MODEL_CONTEXT_SIZE=${SUMMARIZATION_CONTEXT_SIZE:-32768}
      - MODEL_GPU_LAYERS=${SUMMARIZATION_GPU_LAYERS:-999}
      - LLAMA_SERVER_EXTRA_ARGS=${SUMMARIZATION_LLAMA_SERVER_EXTRA_ARGS:-}
      - VLLM_EXTRA_ARGS=${SUMMARIZATION_VLLM_EXTRA_ARGS:-}

  backend:
    build:
      args:
        ROCM_WHL_URL: ${ROCM_WHL_URL:-https://repo.radeon.com/rocm/manylinux/rocm-rel-7.2/}
        ROCM_TORCH_VERSION: ${ROCM_TORCH_VERSION:-2.10.0+rocm7.2.0.lw.gitb6ee5fde}
        ROCM_TORCHAUDIO_VERSION: ${ROCM_TORCHAUDIO_VERSION:-2.10.0+rocm7.2.0.git5047768f}
        ROCM_TORCHVISION_VERSION: ${ROCM_TORCHVISION_VERSION:-0.25.0+rocm7.2.0.git82df5f59}
        ROCM_NIGHTLY: ${ROCM_NIGHTLY:-0}
        LLAMA_BUILD_VULKAN: ${LLAMA_BUILD_VULKAN:-0}
        LLAMA_BUILD_HIPBLAS: ${LLAMA_BUILD_HIPBLAS:-1}
        INSTALL_NEMO: ${INSTALL_NEMO:-0}
    devices:
      - /dev/kfd:/dev/kfd
      - /dev/dri:/dev/dri
    volumes:
      - /opt/rocm:/opt/rocm:ro
    group_add:
      - "44"   # video group
      - "993"  # render group
    environment:
      - GPU_BACKEND=${GPU_BACKEND:-rocm}
      - HARDWARE_PROFILE=strix_halo
      - HSA_OVERRIDE_GFX_VERSION=${HSA_OVERRIDE_GFX_VERSION:-11.5.1}
      - MIOPEN_DEBUG_GCN_ASM_KERNELS=${MIOPEN_DEBUG_GCN_ASM_KERNELS:-0}
      - MODEL_LOADING=${MODEL_LOADING:-sequential}
      - ROCM_LLM_RUNTIME=${ROCM_LLM_RUNTIME:-llama_server}
      - ROCM_LLM_IDLE_TIMEOUT_S=${ROCM_LLM_IDLE_TIMEOUT_S:-120}
      - ROCM_PATH=/opt/rocm
      - LD_LIBRARY_PATH=/opt/rocm/lib:/opt/rocm/lib64
      - VISION_ENDPOINT_URL=http://vision-server:8080/v1
      - VISION_ENDPOINT_MODEL=${VISION_ENDPOINT_MODEL:-Qwen3VL-32B-Instruct-Q4_K_M.gguf}
      - SUMMARIZATION_ENDPOINT_URL=http://summarization-server:8080/v1
      - SUMMARIZATION_ENDPOINT_MODEL=${SUMMARIZATION_ENDPOINT_MODEL:-qwen3-30b-a3b-q4_k_m.gguf}
    depends_on:
      db:
        condition: service_healthy
      redis:
        condition: service_healthy
      vision-server:
        condition: service_healthy
      summarization-server:
        condition: service_healthy

  worker:
    command: celery -A app.workers.celery_app worker --loglevel=info --pool=solo -Q video_processing,batch_processing,default
    build:
      args:
        ROCM_WHL_URL: ${ROCM_WHL_URL:-https://repo.radeon.com/rocm/manylinux/rocm-rel-7.2/}
        ROCM_TORCH_VERSION: ${ROCM_TORCH_VERSION:-2.10.0+rocm7.2.0.lw.gitb6ee5fde}
        ROCM_TORCHAUDIO_VERSION: ${ROCM_TORCHAUDIO_VERSION:-2.10.0+rocm7.2.0.git5047768f}
        ROCM_TORCHVISION_VERSION: ${ROCM_TORCHVISION_VERSION:-0.25.0+rocm7.2.0.git82df5f59}
        ROCM_NIGHTLY: ${ROCM_NIGHTLY:-0}
        LLAMA_BUILD_VULKAN: ${LLAMA_BUILD_VULKAN:-0}
        LLAMA_BUILD_HIPBLAS: ${LLAMA_BUILD_HIPBLAS:-1}
        INSTALL_NEMO: ${INSTALL_NEMO:-0}
    devices:
      - /dev/kfd:/dev/kfd
      - /dev/dri:/dev/dri
    volumes:
      - /opt/rocm:/opt/rocm:ro
    group_add:
      - "44"   # video group
      - "993"  # render group
    environment:
      - GPU_BACKEND=${GPU_BACKEND:-rocm}
      - HARDWARE_PROFILE=strix_halo
      - HSA_OVERRIDE_GFX_VERSION=${HSA_OVERRIDE_GFX_VERSION:-11.5.1}
      - MIOPEN_DEBUG_GCN_ASM_KERNELS=${MIOPEN_DEBUG_GCN_ASM_KERNELS:-0}
      - MODEL_LOADING=${MODEL_LOADING:-sequential}
      - ROCM_LLM_RUNTIME=${ROCM_LLM_RUNTIME:-llama_server}
      - ROCM_LLM_IDLE_TIMEOUT_S=${ROCM_LLM_IDLE_TIMEOUT_S:-120}
      - ROCM_PATH=/opt/rocm
      - LD_LIBRARY_PATH=/opt/rocm/lib:/opt/rocm/lib64
      - VISION_ENDPOINT_URL=http://vision-server:8080/v1
      - VISION_ENDPOINT_MODEL=${VISION_ENDPOINT_MODEL:-Qwen3VL-32B-Instruct-Q4_K_M.gguf}
      - SUMMARIZATION_ENDPOINT_URL=http://summarization-server:8080/v1
      - SUMMARIZATION_ENDPOINT_MODEL=${SUMMARIZATION_ENDPOINT_MODEL:-qwen3-30b-a3b-q4_k_m.gguf}
    depends_on:
      db:
        condition: service_healthy
      redis:
        condition: service_healthy
      vision-server:
        condition: service_healthy
      summarization-server:
        condition: service_healthy

volumes:
  vllm_cache: