Return prompt_tokens via API

[prd-tuong-nguyen]

Feature request

Currently, the API only returned the generated_tokens.
I think the API should return the prompt_tokens as number of token in requested prompt. It really useful to tracking the response speed according to the number of tokens.

Motivation

Use to tracking the speed according the number of token

Your contribution

:(

[tgaddair]

Thanks for raising this issue @prd-tuong-nguyen, we can definitely get this in for the next release!

[tgaddair]

Hey @prd-tuong-nguyen, this is now done following #165. Note that the Python SDK will be updated in PyPI with the next release. Until then, you can install from the repo to get the latest changes.

[prd-tuong-nguyen]

@tgaddair Awesome bro, thank you so much