Is your feature request related to a problem? Please describe.
In order to use the Batch Speech API, the sample code instructs us to generate client code from the Swagger API definition and manually download it and install it via pip.
This introduces an extra manual step for the users, and causes us to either:
store this code in our respective repos (which for the large part will be unused and untested except for the specific APIs we use) and safeguard this code from being edited from with our repos, OR
create a build step where we automate the downloading and installing of this client code, which adds complexity to our builds.
Another issue with this is that we cannot use standard dependency-upgrading tools like dependabot when a new version of the API client is released.
Describe the solution you'd like
We would like the Batch Speech API client for Python to be published as a standard Python package to a repository like PyPi, so that we can use the client in a standard manner just like every other dependency.
Thanks @superhindupur
Just one comment,
We would like to use the python library generated through Swagger for all steps in our Speech Experiment Accelerator (such as training, evaluation, and inference)
Is your feature request related to a problem? Please describe. In order to use the Batch Speech API, the sample code instructs us to generate client code from the Swagger API definition and manually download it and install it via pip.
This introduces an extra manual step for the users, and causes us to either:
Another issue with this is that we cannot use standard dependency-upgrading tools like dependabot when a new version of the API client is released.
Describe the solution you'd like We would like the Batch Speech API client for Python to be published as a standard Python package to a repository like PyPi, so that we can use the client in a standard manner just like every other dependency.
Describe alternatives you've considered NA
Additional context NA