VITS: lang_list is not defined & optional dynamic_loading (#187)

* Update ModelManager.py dynamic_loading default value * Update views.py defined lang_list * Update TTSManager.py 'dynamic_loading' is optional * Update views.py
Artrajz · Jan 5, 2025 · 66ac048 · 66ac048
1 parent 7b4cb9c
commit 66ac048
Show file tree

Hide file tree

Showing 3 changed files with 6 additions and 4 deletions.
diff --git a/manager/ModelManager.py b/manager/ModelManager.py
@@ -247,7 +247,7 @@ def _load_model_from_path(self, tts_model):
                 "device": self.device
             }
             if model_type == ModelType.VITS:
-                model_args["dynamic_loading"] = tts_model["dynamic_loading"]
+                model_args["dynamic_loading"] = tts_model.get("dynamic_loading", False)
 
         model_class = self.model_class_map[model_type]
         model = model_class(**model_args)

diff --git a/manager/TTSManager.py b/manager/TTSManager.py
@@ -271,7 +271,7 @@ def process_ssml_infer_task(self, tasks, format):
 
     def vits_infer(self, state, encode=True):
         model = self.get_model(ModelType.VITS, state["id"])
-        if config.vits_config.dynamic_loading:
+        if hasattr(config.vits_config, 'dynamic_loading') and config.vits_config.dynamic_loading:
             model.load_model()
         state["id"] = self.get_real_id(ModelType.VITS, state["id"])  # Change to real id
         # 去除所有多余的空白字符
@@ -299,7 +299,7 @@ def vits_infer(self, state, encode=True):
                 audios.append(brk)
 
         audio = np.concatenate(audios, axis=0)
-        if config.vits_config.dynamic_loading:
+        if hasattr(config.vits_config, 'dynamic_loading') and config.vits_config.dynamic_loading:
             model.release_model()
         return self.encode(sampling_rate, audio, state["format"]) if encode else audio
 

diff --git a/tts_app/voice_api/views.py b/tts_app/voice_api/views.py
@@ -172,6 +172,8 @@ def voice_vits_api():
     if (lang_detect := config.language_identification.language_automatic_detect) and isinstance(lang_detect, list):
         speaker_lang = lang_detect
 
+    lang_list = get_lang_list(lang, speaker_lang)
+
     if use_streaming and format.upper() != "MP3":
         format = "mp3"
         logger.warning("Streaming response only supports MP3 format.")
@@ -186,7 +188,7 @@ def voice_vits_api():
         "noise": noise,
         "noisew": noisew,
         "segment_size": segment_size,
-        "lang": lang_list,
+        "lang": lang_list[0],
         "speaker_lang": speaker_lang,
     }