Merge pull request #1033 from GOODDAYDAY/feature/doubao-tts

feat: 增加对豆包tts对话速率等控制

main/manager-api/src/main/resources/db/changelog/202504082211.sql

@@ -25,7 +25,7 @@ INSERT INTO `ai_model_provider` (`id`, `model_type`, `provider_code`, `name`, `f
 
 -- TTS模型供应器
 ('SYSTEM_TTS_edge', 'TTS', 'edge', 'Edge TTS', '[{"key":"voice","label":"音色","type":"string"},{"key":"output_dir","label":"输出目录","type":"string"}]', 1, 1, NOW(), 1, NOW()),
-('SYSTEM_TTS_doubao', 'TTS', 'doubao', '火山引擎TTS', '[{"key":"api_url","label":"API地址","type":"string"},{"key":"voice","label":"音色","type":"string"},{"key":"output_dir","label":"输出目录","type":"string"},{"key":"authorization","label":"授权","type":"string"},{"key":"appid","label":"应用ID","type":"string"},{"key":"access_token","label":"访问令牌","type":"string"},{"key":"cluster","label":"集群","type":"string"}]', 2, 1, NOW(), 1, NOW()),
+('SYSTEM_TTS_doubao', 'TTS', 'doubao', '火山引擎TTS', '[{"key": "api_url","label": "API地址","type": "string"},{"key": "voice","label": "音色","type": "string"},{"key": "output_dir","label": "输出目录","type": "string"},{"key": "authorization","label": "授权","type": "string"},{"key": "appid","label": "应用ID","type": "string"},{"key": "access_token","label": "访问令牌","type": "string"},{"key": "cluster","label": "集群","type": "string"},{"key": "speed_ratio","label": "语速","type": "number"},{"key": "volume_ratio","label": "音量","type": "number"},{"key": "pitch_ratio","label": "音高","type": "number"}]', 2, 1, NOW(), 1, NOW()),
 ('SYSTEM_TTS_siliconflow', 'TTS', 'siliconflow', '硅基流动TTS', '[{"key":"model","label":"模型","type":"string"},{"key":"voice","label":"音色","type":"string"},{"key":"output_dir","label":"输出目录","type":"string"},{"key":"access_token","label":"访问令牌","type":"string"},{"key":"response_format","label":"响应格式","type":"string"}]', 3, 1, NOW(), 1, NOW()),
 ('SYSTEM_TTS_cozecn', 'TTS', 'cozecn', 'COZECN TTS', '[{"key":"voice","label":"音色","type":"string"},{"key":"output_dir","label":"输出目录","type":"string"},{"key":"access_token","label":"访问令牌","type":"string"},{"key":"response_format","label":"响应格式","type":"string"}]', 4, 1, NOW(), 1, NOW()),
 ('SYSTEM_TTS_fishspeech', 'TTS', 'fishspeech', 'FishSpeech TTS', '[{"key":"output_dir","label":"输出目录","type":"string"},{"key":"response_format","label":"响应格式","type":"string"},{"key":"reference_id","label":"参考ID","type":"string"},{"key":"reference_audio","label":"参考音频","type":"dict","dict_name":"reference_audio"},{"key":"reference_text","label":"参考文本","type":"dict","dict_name":"reference_text"},{"key":"normalize","label":"是否标准化","type":"boolean"},{"key":"max_new_tokens","label":"最大新令牌数","type":"number"},{"key":"chunk_length","label":"块长度","type":"number"},{"key":"top_p","label":"top_p值","type":"number"},{"key":"repetition_penalty","label":"重复惩罚","type":"number"},{"key":"temperature","label":"温度","type":"number"},{"key":"streaming","label":"是否流式","type":"boolean"},{"key":"use_memory_cache","label":"是否使用内存缓存","type":"string"},{"key":"seed","label":"种子","type":"number"},{"key":"channels","label":"通道数","type":"number"},{"key":"rate","label":"采样率","type":"number"},{"key":"api_key","label":"API密钥","type":"string"},{"key":"api_url","label":"API地址","type":"string"}]', 5, 1, NOW(), 1, NOW()),

main/xiaozhi-server/config.yaml

@@ -397,6 +397,9 @@ TTS:
     appid: 你的火山引擎语音合成服务appid
     access_token: 你的火山引擎语音合成服务access_token
     cluster: volcano_tts
+    speed_ratio: 1.0
+    volume_ratio: 1.0
+    pitch_ratio: 1.0
   CosyVoiceSiliconflow:
     type: siliconflow
     # 硅基流动TTS

main/xiaozhi-server/core/providers/tts/doubao.py

@@ -27,6 +27,10 @@ class TTSProvider(TTSProviderBase):
         else:
             self.voice = config.get("voice")
 
+        self.speed_ratio = float(config.get("speed_ratio", 0.1))
+        self.volume_ratio = float(config.get("volume_ratio", 0.1))
+        self.pitch_ratio = float(config.get("pitch_ratio", 0.1))
+
         self.api_url = config.get("api_url")
         self.authorization = config.get("authorization")
         self.header = {"Authorization": f"{self.authorization}{self.access_token}"}
@@ -49,9 +53,9 @@ class TTSProvider(TTSProviderBase):
             "audio": {
                 "voice_type": self.voice,
                 "encoding": "wav",
-                "speed_ratio": 1.0,
+                "speed_ratio": self.speed_ratio,
-                "volume_ratio": 1.0,
+                "volume_ratio": self.volume_ratio,
-                "pitch_ratio": 1.0,
+                "pitch_ratio": self.pitch_ratio,
             },
             "request": {
                 "reqid": str(uuid.uuid4()),