This example project demonstrates how to use Amazon Transcribe in real-time with Python and Websockets. It contains no dependencies on Boto3.

• Python 3.10
• websockets (pip install websockets)
• AWS credentials

The example file is named example.py. It will send audio in 100ms 'chunks' per payload, serialized/marshalled into the AWS EventStream AudioEvent format. The file that gets streamed is example_call_2_channel.wav. The JSON response from Amazon Transcribe is parsed and the partial result transcripts are printed to the terminal.

Before running the example, make sure that you have your AWS AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY, and AWS_DEFAULT_REGION (region is optional, and will default to us-east-1). You can fill them into the variables defined in example.py, or set environment variables with the same names. Please note: If you are using temporary credentials, for example, from STS, you will also need to set the AWS_SESSION_TOKEN.

If you are swapping out the audio file for something else, please configure at the # Sound settings variables and update the media_encoding, number_of_channels,

To run the example, navigate via the cli to the folder, and type python3 example.py

See CONTRIBUTING for more information.

This library is licensed under the MIT-0 License. See the LICENSE file.