fix-no-f0-model-protect-issue

fix-no-f0-model-protect-issue

Author: RVC-Boss

infer-web.py

@@ -3,7 +3,7 @@
 import sys
 now_dir = os.getcwd()
 sys.path.append(now_dir)
-import traceback
+import traceback,pdb
 import warnings
 
 import numpy as np
@@ -396,7 +396,7 @@ def uvr(model_name, inp_root, save_root_vocal, paths, save_root_ins, agg, format
 
 
 # 一个选项卡全局只能有一个音色
-def get_vc(sid):
+def get_vc(sid,to_return_protect0,to_return_protect1):
     global n_spk, tgt_sr, net_g, vc, cpt, version
     if sid == "" or sid == []:
         global hubert_model
@@ -434,6 +434,11 @@ def get_vc(sid):
     tgt_sr = cpt["config"][-1]
     cpt["config"][-3] = cpt["weight"]["emb_g.weight"].shape[0]  # n_spk
     if_f0 = cpt.get("f0", 1)
+    if(if_f0==0):
+        to_return_protect0=to_return_protect1={"visible": False, "value": 0.5, "__type__": "update"}
+    else:
+        to_return_protect0 ={"visible": True, "value": to_return_protect0, "__type__": "update"}
+        to_return_protect1 ={"visible": True, "value": to_return_protect1, "__type__": "update"}
     version = cpt.get("version", "v1")
     if version == "v1":
         if if_f0 == 1:
@@ -454,7 +459,7 @@ def get_vc(sid):
         net_g = net_g.float()
     vc = VC(tgt_sr, config)
     n_spk = cpt["config"][-3]
-    return {"visible": True, "maximum": n_spk, "__type__": "update"}
+    return {"visible": True, "maximum": n_spk, "__type__": "update"},to_return_protect0,to_return_protect1
 
 
 def change_choices():
@@ -1247,11 +1252,6 @@ def export_onnx(ModelPath, ExportedPath):
                     interactive=True,
                 )
                 clean_button.click(fn=clean, inputs=[], outputs=[sid0])
-                sid0.change(
-                    fn=get_vc,
-                    inputs=[sid0],
-                    outputs=[spk_item],
-                )
             with gr.Group():
                 gr.Markdown(
                     value=i18n("男转女推荐+12key, 女转男推荐-12key, 如果音域爆炸导致音色失真也可以自己调整到合适音域. ")
@@ -1475,6 +1475,11 @@ def export_onnx(ModelPath, ExportedPath):
                         ],
                         [vc_output3],
                     )
+            sid0.change(
+                fn=get_vc,
+                inputs=[sid0,protect0,protect1],
+                outputs=[spk_item,protect0,protect1],
+            )
         with gr.TabItem(i18n("伴奏人声分离&去混响&去回声")):
             with gr.Group():
                 gr.Markdown(

vc_infer_pipeline.py

@@ -184,7 +184,7 @@ def vc(
         with torch.no_grad():
             logits = model.extract_features(**inputs)
             feats = model.final_proj(logits[0]) if version == "v1" else logits[0]
-        if protect < 0.5:
+        if protect < 0.5 and pitch!=None and pitchf!=None:
             feats0 = feats.clone()
         if (
             isinstance(index, type(None)) == False
@@ -211,7 +211,7 @@ def vc(
             )
 
         feats = F.interpolate(feats.permute(0, 2, 1), scale_factor=2).permute(0, 2, 1)
-        if protect < 0.5:
+        if protect < 0.5 and pitch!=None and pitchf!=None:
             feats0 = F.interpolate(feats0.permute(0, 2, 1), scale_factor=2).permute(
                 0, 2, 1
             )
@@ -223,7 +223,7 @@ def vc(
                 pitch = pitch[:, :p_len]
                 pitchf = pitchf[:, :p_len]
 
-        if protect < 0.5:
+        if protect < 0.5 and pitch!=None and pitchf!=None:
             pitchff = pitchf.clone()
             pitchff[pitchf > 0] = 1
             pitchff[pitchf < 1] = protect