ezlocalai/docker-compose-local.yml at main · DevXT-LLC/ezlocalai · 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
services:
  ezlocalai:
    build:
      context: .
      dockerfile: Dockerfile
    env_file:
      - .env
    environment:
      - EZLOCALAI_URL=${EZLOCALAI_URL:-http://localhost:8091}
      - EZLOCALAI_API_KEY=${EZLOCALAI_API_KEY:-}
      - DEFAULT_MODEL=${DEFAULT_MODEL:-unsloth/Qwen3.5-4B-GGUF}
      - WHISPER_MODEL=${WHISPER_MODEL:-large-v3}
      - IMG_MODEL=${IMG_MODEL:-}
      - VIDEO_MODEL=${VIDEO_MODEL:-}
      - LLM_BATCH_SIZE=${LLM_BATCH_SIZE:-1024}
      - MAX_CONCURRENT_REQUESTS=${MAX_CONCURRENT_REQUESTS:-1}
      - MAX_QUEUE_SIZE=${MAX_QUEUE_SIZE:-100}
      - REQUEST_TIMEOUT=${REQUEST_TIMEOUT:-300}
      - VOICE_SERVER=${VOICE_SERVER:-}
      - VOICE_SERVER_API_KEY=${VOICE_SERVER_API_KEY:-}
      - LAZY_LOAD_VOICE=${LAZY_LOAD_VOICE:-true}
    restart: unless-stopped
    healthcheck:
      test: ["CMD", "curl", "-sf", "http://localhost:8091/health"]
      interval: 30s
      timeout: 10s
      retries: 3
      start_period: 120s
    logging:
      driver: json-file
      options:
        max-size: "500m"
        max-file: "5"
    ulimits:
      nofile:
        soft: 65535
        hard: 65535
    ports:
      - "8091:8091"
    volumes:
      - ./models:/app/models
      - ./hf:/home/root/.cache/huggingface/hub
      - ./outputs:/app/outputs
      - ./voices:/app/voices
      - ./whispercpp:/app/whispercpp