feat: update min_speaker_count and max_speaker_count to be optional
docs: Clarify that min_speaker_count and max_speaker_count in SpeakerDiarizationConfig are not currently used
docs: Update guidance on how to enable speaker diarization; to enable, set the diarization_config field to an empty SpeakerDiarizationConfig message
PiperOrigin-RevId: 852383212
docs: update the list of automatically detected encodings
docs: update the maximum number of files that may be included in a BatchRecognizeRequest
PiperOrigin-RevId: 653986958
docs: clarify alternatives for deprecated fields
feat: deprecate `BatchRecognizeFileResult.uri` in favor of `cloud_storage_result.native_format_uri`
feat: deprecate `BatchRecognizeFileResult.transcript` in favor of `inline_result.transcript`
PiperOrigin-RevId: 577926708
Enables specifying `model` and `language_codes` in requests without having to specify them in the Recognizer (they can still be specified in the Recognizer in the `default_recognition_config` field).
Also enables using the recognizer ID `_` to perform recognition without explicitly creating a Recognizer resource.
The top-level `model` and `language_codes` fields are deprecated in favor of the new fields added in the `RecognitionConfig` message. The old fields continue to work.
PiperOrigin-RevId: 545698919
The next generation Speech-to-Text V2 API provides the same features clients are familiar with in the existing Speech-to-Text API, while modernizing and simplifying the API using API best practices, resourceful methods, and updated infrastructure. Although this proto is published, the service is not yet available.
PiperOrigin-RevId: 474926510
