refactor: api

Author: luoling8192

apps/silero-vad-whisper-realtime-api/src/api.rs

@@ -1,6 +1,4 @@
-
-use serde::Serialize;
-
+use serde::{Deserialize, Serialize};
 
 #[derive(Debug, Serialize)]
 pub struct TranscriptionResponse {
@@ -27,3 +25,43 @@ pub struct ErrorDetail {
   pub param: Option<String>,
   pub code: Option<String>,
 }
+
+pub fn default_model() -> String {
+  "whisper-1".to_string()
+}
+
+pub fn default_response_format() -> String {
+  "json".to_string()
+}
+
+pub fn default_temperature() -> f32 {
+  0.0
+}
+
+#[derive(Debug, Deserialize)]
+pub struct TranscriptionRequest {
+  /// The audio file object to transcribe
+  /// In multipart form, this would be the file field
+
+  /// ID of the model to use. Only whisper-1 is currently available.
+  #[serde(default = "default_model")]
+  pub model: String,
+
+  /// The language of the input audio
+  pub language: Option<String>,
+
+  /// An optional text to guide the model's style or continue a previous audio segment
+  pub prompt: Option<String>,
+
+  /// The format of the transcript output
+  #[serde(default = "default_response_format")]
+  pub response_format: String,
+
+  /// The sampling temperature, between 0 and 1
+  #[serde(default = "default_temperature")]
+  pub temperature: f32,
+
+  /// Enable streaming response
+  #[serde(default)]
+  pub stream: bool,
+}

apps/silero-vad-whisper-realtime-api/src/asr.rs

@@ -1,7 +1,7 @@
 use std::{collections::HashMap, sync::Arc};
 
-use crate::api::{ErrorDetail, ErrorResponse, StreamChunk, TranscriptionResponse};
-use crate::{AppState, TranscriptionRequest, default_model, default_response_format, default_temperature};
+use crate::AppState;
+use crate::api::{default_model, default_response_format, default_temperature, ErrorDetail, ErrorResponse, StreamChunk, TranscriptionRequest, TranscriptionResponse};
 
 use crate::audio_manager::AudioBuffer;
 use anyhow::Result;

apps/silero-vad-whisper-realtime-api/src/main.rs

@@ -7,63 +7,21 @@ use axum::{
   routing::{get, post},
 };
 use candle_core::Device;
-use serde::Deserialize;
 use tokio::sync::Mutex;
 use tower::ServiceBuilder;
 use tower_http::cors::CorsLayer;
 
 use crate::{
-  asr::{transcribe_audio},
+  asr::transcribe_audio,
   vad::VADProcessor,
   whisper::{WhichWhisperModel, WhisperProcessor},
 };
 
+mod api;
+mod asr;
 mod audio_manager;
 mod vad;
 mod whisper;
-mod asr;
-mod api;
-
-#[derive(Debug, Deserialize)]
-pub struct TranscriptionRequest {
-  /// The audio file object to transcribe
-  /// In multipart form, this would be the file field
-
-  /// ID of the model to use. Only whisper-1 is currently available.
-  #[serde(default = "default_model")]
-  model: String,
-
-  /// The language of the input audio
-  language: Option<String>,
-
-  /// An optional text to guide the model's style or continue a previous audio segment
-  prompt: Option<String>,
-
-  /// The format of the transcript output
-  #[serde(default = "default_response_format")]
-  response_format: String,
-
-  /// The sampling temperature, between 0 and 1
-  #[serde(default = "default_temperature")]
-  temperature: f32,
-
-  /// Enable streaming response
-  #[serde(default)]
-  stream: bool,
-}
-
-fn default_model() -> String {
-  "whisper-1".to_string()
-}
-
-fn default_response_format() -> String {
-  "json".to_string()
-}
-
-fn default_temperature() -> f32 {
-  0.0
-}
-
 
 // Application state
 struct AppState {