openCHS

Appearance

Whisper Model Documentation ​

1. Model Overview ​

The Whisper model is a state-of-the-art speech recognition model from OpenAI. It is designed to transcribe audio from a variety of languages and accents with high accuracy.

2. Model Details ​

  • Model: The model is the whisper-large-v3-turbo model from OpenAI.
  • Languages: The model supports over 99 languages.

3. API Endpoints ​

3.1. Transcribe Audio ​

  • Method: POST
  • Path: /whisper/transcribe
  • Description: Transcribes an audio file to text.
  • Request Body: multipart/form-data with an audio file.
  • Response: A JSON object with the transcript.

4. Request Format ​

The request body for the /whisper/transcribe endpoint should be a multipart/form-data request with the following field:

  • audio: The audio file to be transcribed.

5. Response Format ​

The response from the /whisper/transcribe endpoint is a JSON object with the following fields:

json
{
  "transcript": "The transcribed text.",
  "language": "The detected language of the audio.",
  "processing_time": "The time taken to process the request.",
  "model_info": {
    "model_path": "The path to the model.",
    "loaded": "Whether the model is loaded.",
    "load_time": "The time when the model was loaded.",
    "device": "The device on which the model is running.",
    "error": "Any error that occurred during model loading."
  },
  "timestamp": "The timestamp of the request.",
  "audio_info": {
    "filename": "The name of the audio file.",
    "file_size_mb": "The size of the audio file in megabytes.",
    "format": "The format of the audio file.",
    "content_type": "The content type of the audio file."
  }
}

6. Example Usage ​

bash
curl -X POST \
  -F "audio=@/path/to/your/audio.wav" \
  http://localhost:8123/whisper/transcribe

7. Error Handling ​

If the Whisper model is not loaded, the API will return a 503 Service Unavailable error. If the request is invalid, the API will return a 400 Bad Request error.