修复 #2354

修改`api.py`使其可以正常兼容v4模型。已经过测试，可以使用。

api.py

@@ -176,9 +176,9 @@ import subprocess
 
 class DefaultRefer:
     def __init__(self, path, text, language):
-        self.path = args.default_refer_path
-        self.text = args.default_refer_text
-        self.language = args.default_refer_language
+        self.path = path
+        self.text = text
+        self.language = language
 
     def is_ready(self) -> bool:
         return is_full(self.path, self.text, self.language)
@@ -294,11 +294,13 @@ from process_ckpt import get_sovits_version_from_path_fast, load_sovits_new
 
 def get_sovits_weights(sovits_path):
     path_sovits_v3 = "GPT_SoVITS/pretrained_models/s2Gv3.pth"
-    is_exist_s2gv3 = os.path.exists(path_sovits_v3)
+    path_sovits_v4 = "GPT_SoVITS/pretrained_models/gsv-v4-pretrained/s2Gv4.pth"
 
     version, model_version, if_lora_v3 = get_sovits_version_from_path_fast(sovits_path)
-    if if_lora_v3 == True and is_exist_s2gv3 == False:
+    if if_lora_v3 == True and not os.path.exists(path_sovits_v3):
         logger.info("SoVITS V3 底模缺失，无法加载相应 LoRA 权重")
+    if model_version == "v4" and not os.path.exists(path_sovits_v4):
+        logger.info("SoVITS V4 底模缺失，无法加载相应 LoRA 权重")
 
     dict_s2 = load_sovits_new(sovits_path)
     hps = dict_s2["config"]
@@ -310,12 +312,13 @@ def get_sovits_weights(sovits_path):
         hps.model.version = "v1"
     else:
         hps.model.version = "v2"
-
     if model_version == "v3":
         hps.model.version = "v3"
+    if model_version == "v4":
+        hps.model.version = "v4"
 
     model_params_dict = vars(hps.model)
-    if model_version != "v3":
+    if model_version != "v3" and model_version != "v4":
         vq_model = SynthesizerTrn(
             hps.data.filter_length // 2 + 1,
             hps.train.segment_size // hps.data.hop_length,
@@ -342,10 +345,13 @@ def get_sovits_weights(sovits_path):
     else:
         vq_model = vq_model.to(device)
     vq_model.eval()
-    if if_lora_v3 == False:
+    if if_lora_v3 == False or model_version != "v4":
         vq_model.load_state_dict(dict_s2["weight"], strict=False)
     else:
-        vq_model.load_state_dict(load_sovits_new(path_sovits_v3)["weight"], strict=False)
+        if model_version == "v4":
+            vq_model.load_state_dict(load_sovits_new(path_sovits_v4)["weight"], strict=False)
+        else:
+            vq_model.load_state_dict(load_sovits_new(path_sovits_v3)["weight"], strict=False)
         lora_rank = dict_s2["lora_rank"]
         lora_config = LoraConfig(
             target_modules=["to_k", "to_q", "to_v", "to_out.0"],
@@ -394,13 +400,11 @@ def change_gpt_sovits_weights(gpt_path, sovits_path):
     try:
         gpt = get_gpt_weights(gpt_path)
         sovits = get_sovits_weights(sovits_path)
+        speaker_list["default"] = Speaker(name="default", gpt=gpt, sovits=sovits)
+        return JSONResponse({"code": 0, "message": "Success"}, status_code=200)
     except Exception as e:
         return JSONResponse({"code": 400, "message": str(e)}, status_code=400)
     
-    speaker_list["default"] = Speaker(name="default", gpt=gpt, sovits=sovits)
-    return JSONResponse({"code": 0, "message": "Success"}, status_code=200)
-
-
 def get_bert_feature(text, word2ph):
     with torch.no_grad():
         inputs = tokenizer(text, return_tensors="pt")
@@ -759,7 +763,7 @@ def get_tts_wav(
         prompt_semantic = codes[0, 0]
         prompt = prompt_semantic.unsqueeze(0).to(device)
 
-        if version != "v3":
+        if version != "v3" and version != "v4":
             refers = []
             if inp_refs:
                 for path in inp_refs:
@@ -811,7 +815,7 @@ def get_tts_wav(
             pred_semantic = pred_semantic[:, -idx:].unsqueeze(0)
         t3 = ttime()
 
-        if version != "v3":
+        if version != "v3" and version != "v4":
             audio = (
                 vq_model.decode(pred_semantic, torch.LongTensor(phones2).to(device).unsqueeze(0), refers, speed=speed)
                 .detach()
@@ -880,7 +884,7 @@ def get_tts_wav(
         audio_opt = np.concatenate(audio_opt, 0)
         t4 = ttime()
 
-        sr = hps.data.sampling_rate if version != "v3" else 24000
+        sr = hps.data.sampling_rate if version != "v3" and version != "v4" else 24000
         if if_sr and sr == 24000:
             audio_opt = torch.from_numpy(audio_opt).float().to(device)
             audio_opt, sr = audio_sr(audio_opt.unsqueeze(0), sr)
@@ -901,7 +905,7 @@ def get_tts_wav(
     if not stream_mode == "normal":
         if media_type == "wav":
             sr = 48000 if if_sr else 24000
-            sr = hps.data.sampling_rate if version != "v3" else sr
+            sr = hps.data.sampling_rate if version != "v3" and version != "v4" else sr
             audio_bytes = pack_wav(audio_bytes, sr)
         yield audio_bytes.getvalue()