protos/text_to_speech.proto

syntax = "proto3";

package texttospeech;

service TextToSpeechService {
  rpc ProcessText (stream ProcessTextRequest) returns (stream ProcessTextResponse);
}

message ProcessTextRequest {
  oneof request_data {
    Audio audio_data = 1;
    Meta metadata = 2;
    Status status = 3;
  }
}

message ProcessTextResponse {
  bytes buffer = 1;
  string session_id = 2;
  string sequence_id = 3;
  string transcript = 4;
}

message Meta {
  string session_id = 1;
  int32 silenceDuration = 2;
  int32 threshold = 3;
  float temperature = 4;
  string activeVoice = 5;
  int32 maxTokens = 6;
}

message Status {
  string transcript = 1;
  string played_seq = 2;
  string interrupt_seq = 3;
}

message Audio {
  bytes buffer = 1;
}