tabby/deployment/docker-compose.yml

version: '3.3'

services:
  server:
    image: tabbyml/tabby
    container_name: tabby-server
    working_dir: /app/server
    command: uvicorn app:app --host 0.0.0.0 --port 5000
    environment:
      - TOKENIZER_NAME=/tokenizer
      - TRITON_HOST=triton
      - EVENTS_LOG_DIR=/logs/tabby-server
    ports:
      - "5000:5000"
    volumes:
      - ./logs:/logs
      - ../testdata/gptneox/tokenizer:/tokenizer
    links:
      - triton
      - vector

  triton:
    image: tabbyml/fastertransformer_backend
    container_name: tabby-triton
    command: mpirun -n 1 --allow-run-as-root /opt/tritonserver/bin/tritonserver --model-repository=/model
    shm_size: 1gb
    volumes:
      - ../testdata/gptneox/models:/model
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: all
              capabilities: [gpu]

  admin:
    image: tabbyml/tabby
    container_name: tabby-admin
    working_dir: /app/admin
    command: streamlit run Home.py
    ports:
      - "8501:8501"
    links:
      - server
      - triton
      - vector

  vector:
    image: timberio/vector:0.28.1-alpine
    container_name: tabby-vector
    volumes:
      - ./config/vector.toml:/etc/vector/vector.toml:ro
      - ./logs:/logs