How to Use Nova Sonic Speech-to-Speech with Python (boto3 or aws_sdk_bedrock_runtime?)

[yourappleintelligence]

I’m trying to use the speech-to-speech (Python example), but I keep encountering errors related to:

ERROR: No matching distribution found for aws_sdk_bedrock_runtime

I haven’t been able to find a corresponding pip package for aws_sdk_bedrock_runtime.

How can I install or use it?

I also found this GitHub repo: https://github.com/awslabs/aws-sdk-python, but I’m not sure how to use it as a package in my project. Can you guide me on that?

Additionally, I want to know: Can I use Nova Sonic (https://docs.aws.amazon.com/nova/latest/userguide/speech-bidirection.html) with boto3, or do I have to use aws_sdk_bedrock_runtime to work with it? because code snippets example in nova sonic docs I can only see aws_sdk_bedrock_runtime usage.

[lanazhang]

Have you tried upgrading pip before installing? For example:
pip install --upgrade pip
If the issue persists after the upgrade, could you let me know your operating system and Python version?

[Roline-Stapny]

Have you tried installing dependencies present in requirements.txt

pyaudio>=0.2.13
rx>=3.2.0
smithy-aws-core>=0.0.1
pytz
aws_sdk_bedrock_runtime

pip install -r requirements.txt

You might want to check if Python version is >= 3.12

[yourappleintelligence]

Thanks @lanazhang and @Roline-Stapny it resolved. issue was due to the python version.