Format code (#330)

Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>

config.py

@@ -2,6 +2,7 @@ import argparse
 import torch
 from multiprocessing import cpu_count
 
+
 def config_file_change_fp32():
     for config_file in ["32k.json", "40k.json", "48k.json"]:
         with open(f"configs/{config_file}", "r") as f:
@@ -13,6 +14,7 @@ def config_file_change_fp32():
     with open("trainset_preprocess_pipeline_print.py", "w") as f:
         f.write(strr)
 
+
 class Config:
     def __init__(self):
         self.device = "cuda:0"

export_onnx.py

@@ -52,4 +52,3 @@ if __name__ == "__main__":
         input_names=input_names,
         output_names=output_names,
     )
-

gui.py

@@ -1,4 +1,4 @@
-'''
+"""
 0416后的更新：
     引入config中half
     重建npy而不用填写
@@ -9,12 +9,14 @@
     int16：
     增加无索引支持
     f0算法改harvest(怎么看就只有这个会影响CPU占用)，但是不这么改效果不好
-'''
+"""
 import os, sys, traceback
+
 now_dir = os.getcwd()
 sys.path.append(now_dir)
 from config import Config
-is_half=Config().is_half
+
+is_half = Config().is_half
 import PySimpleGUI as sg
 import sounddevice as sd
 import noisereduce as nr
@@ -26,7 +28,12 @@ import torchaudio.transforms as tat
 import scipy.signal as signal
 
 # import matplotlib.pyplot as plt
-from infer_pack.models import SynthesizerTrnMs256NSFsid, SynthesizerTrnMs256NSFsid_nono,SynthesizerTrnMs768NSFsid,SynthesizerTrnMs768NSFsid_nono
+from infer_pack.models import (
+    SynthesizerTrnMs256NSFsid,
+    SynthesizerTrnMs256NSFsid_nono,
+    SynthesizerTrnMs768NSFsid,
+    SynthesizerTrnMs768NSFsid_nono,
+)
 from i18n import I18nAuto
 
 i18n = I18nAuto()
@@ -63,7 +70,7 @@ class RVC:
             )
             self.model = models[0]
             self.model = self.model.to(device)
-            if(is_half==True):
+            if is_half == True:
                 self.model = self.model.half()
             else:
                 self.model = self.model.float()
@@ -75,21 +82,25 @@ class RVC:
             self.version = cpt.get("version", "v1")
             if version == "v1":
                 if if_f0 == 1:
-                    self.net_g = SynthesizerTrnMs256NSFsid(*cpt["config"], is_half=config.is_half)
+                    self.net_g = SynthesizerTrnMs256NSFsid(
+                        *cpt["config"], is_half=config.is_half
+                    )
                 else:
                     self.net_g = SynthesizerTrnMs256NSFsid_nono(*cpt["config"])
             elif version == "v2":
                 if if_f0 == 1:
-                    self.net_g = SynthesizerTrnMs768NSFsid(*cpt["config"], is_half=config.is_half)
+                    self.net_g = SynthesizerTrnMs768NSFsid(
+                        *cpt["config"], is_half=config.is_half
+                    )
                 else:
                     self.net_g = SynthesizerTrnMs768NSFsid_nono(*cpt["config"])
             del self.net_g.enc_q
             print(self.net_g.load_state_dict(cpt["weight"], strict=False))
             self.net_g.eval().to(device)
-            if(is_half==True):
+            if is_half == True:
-                self.net_g=self.net_g.half()
+                self.net_g = self.net_g.half()
             else:
-                self.net_g=self.net_g.float()
+                self.net_g = self.net_g.float()
         except:
             print(traceback.format_exc())
 
@@ -151,15 +162,18 @@ class RVC:
 
         ####索引优化
         try:
-            if hasattr(self, "index") and hasattr(self, "big_npy") and self.index_rate != 0:
+            if (
+                hasattr(self, "index")
+                and hasattr(self, "big_npy")
+                and self.index_rate != 0
+            ):
                 npy = feats[0].cpu().numpy().astype("float32")
                 score, ix = self.index.search(npy, k=8)
                 weight = np.square(1 / score)
                 weight /= weight.sum(axis=1, keepdims=True)
-                npy = np.sum(
+                npy = np.sum(self.big_npy[ix] * np.expand_dims(weight, axis=2), axis=1)
-                    self.big_npy[ix] * np.expand_dims(weight, axis=2), axis=1
+                if is_half == True:
-                )
+                    npy = npy.astype("float16")
-                if(is_half==True):npy=npy.astype("float16")
                 feats = (
                     torch.from_numpy(npy).unsqueeze(0).to(device) * self.index_rate
                     + (1 - self.index_rate) * feats

i18n.py

@@ -18,11 +18,11 @@ class I18nAuto:
         if not os.path.exists(f"./i18n/{language}.json"):
             language = "en_US"
         self.language = language
-        #print("Use Language:", language)
+        # print("Use Language:", language)
         self.language_map = load_language_list(language)
 
     def __call__(self, key):
         return self.language_map.get(key, key)
-    
+
     def print(self):
         print("Use Language:", self.language)

infer-web.py

@@ -1069,6 +1069,8 @@ def change_info_(ckpt_path):
 
 
 from infer_pack.models_onnx import SynthesizerTrnMsNSFsidM
+
+
 def export_onnx(ModelPath, ExportedPath, MoeVS=True):
     cpt = torch.load(ModelPath, map_location="cpu")
     cpt["config"][-3] = cpt["weight"]["emb_g.weight"].shape[0]  # n_spk

infer_pack/models_onnx.py

@@ -633,14 +633,14 @@ class SynthesizerTrnMsNSFsidM(nn.Module):
         self.speaker_map = self.speaker_map.unsqueeze(0)
 
     def forward(self, phone, phone_lengths, pitch, nsff0, g, rnd, max_len=None):
-        if self.speaker_map is not None:   # [N, S]  *  [S, B, 1, H]
+        if self.speaker_map is not None:  # [N, S]  *  [S, B, 1, H]
             g = g.reshape((g.shape[0], g.shape[1], 1, 1, 1))  # [N, S, B, 1, 1]
             g = g * self.speaker_map  # [N, S, B, 1, H]
-            g = torch.sum(g, dim=1) # [N, 1, B, 1, H]
+            g = torch.sum(g, dim=1)  # [N, 1, B, 1, H]
-            g = g.transpose(0, -1).transpose(0, -2).squeeze(0) # [B, H, N]
+            g = g.transpose(0, -1).transpose(0, -2).squeeze(0)  # [B, H, N]
         else:
             g = g.unsqueeze(0)
-            g = self.emb_g(g).transpose(1,2)
+            g = self.emb_g(g).transpose(1, 2)
 
         m_p, logs_p, x_mask = self.enc_p(phone, pitch, phone_lengths)
         z_p = (m_p + torch.exp(logs_p) * rnd) * x_mask