Audio Transcriber API

Transcribe audio from URL or upload from local disk.

Running on develoment machine

Make sure you have Python and Redis installed on your system. After cloning this repo, follow these steps.

1. Create and activate virtual env:

   python -m venv venv
   . venv/bin/activate
   

2. Install the dependencies:

   pip install -r requirements.txt 
   

3. Copy the .env_sample file into .env and modify its values:

   REDIS_HOST=localhost
   REDIS_PORT=6379
   

4. Run the development server:

   uvicorn src.main:app --reload
   

   It will run on localhost and accepting requests on port 8000.

5. Open another terminal and activate the virtual env as well:

   . venv/bin/activate
   

6. Run the celery task:

   celery -A src.tasks --loglevel=info
   

Once the API is up and running, you can transcribe your audio files e.g:

import time
import requests 

API_URL = 'http://localhost:8000'

# Upload audio file
with open('/path/to/audio.wav', 'rb') as f:
    r = requests.post(f'{API_URL}/transcribe', files={'audio': f})
    r.raise_for_status()

# Check the status and retrieve the transcriptin
task_id = r.json()['taskId']
while True:
    r = requests.get(f'{API_URL}/transcribe/{task_id}')
    status = r.json()['status']
    print(f'Status = {status}')
    if status == 'DONE':
        break
    time.sleep(1)

print('Result = ', r.json()['result'])

Run with Docker

If you have Docker installed on your machine, you can run the API using Docker:

1. Copy the .env_sample file into .env and modify its values:

   REDIS_HOST=redis
   REDIS_PORT=6379
   

2. Run the app:

   docker compose up
   

   The API will be running and accepting requests on port 8000.

Documentation

FastAPI automatically generates the documentation for the API, you can access the docs at http://localhost:8000/docs.

License

MIT