fortemi/docker-compose.whisper.yml at main · Fortemi/fortemi · 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
# docker-compose.whisper.yml
# Speaches (faster-whisper-server) — OpenAI-compatible audio transcription
# Deploy alongside the main fortemi bundle for audio extraction support.
#
# Usage:
#   docker compose -f docker-compose.whisper.yml up -d
#
# Test:
#   curl http://localhost:8000/health
#   curl http://localhost:8000/v1/audio/transcriptions -F "file=@audio.mp3"
#
# Configure fortemi to use this service:
#   WHISPER_BASE_URL=http://host.docker.internal:8000
#   (or http://speaches:8000 if on same Docker network)

services:
  speaches:
    container_name: speaches
    image: ghcr.io/speaches-ai/speaches:latest-cuda-12.6.3
    restart: unless-stopped
    ports:
      - "8000:8000"
    volumes:
      - speaches-models:/home/ubuntu/.cache/huggingface/hub
    environment:
      - WHISPER__MODEL=${WHISPER_MODEL:-Systran/faster-distil-whisper-large-v3}
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: 1
              capabilities: [gpu]
    healthcheck:
      test: ["CMD", "curl", "--fail", "http://0.0.0.0:8000/health"]
      interval: 30s
      timeout: 10s
      retries: 3
      start_period: 60s

volumes:
  speaches-models:
    driver: local