whisper-large (float32)

docker-compose.yml

version: "3.9"

services:

  whisper:
    image: onerahmet/openai-whisper-asr-webservice:latest-gpu
    restart: "unless-stopped"
    environment:
      - NVIDIA_VISIBLE_DEVICES=all
      - ASR_MODEL=large-v2
    volumes:
      - ./whisper_data:/root/.cache/whisper
      - ./huggingface_data:/root/.cache/huggingface
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: 1
              capabilities: [ gpu ]
    ports:
     - 9000:9000
    logging:
      driver: "json-file"
      options:
        max-size: "50k"