Refactor config (#185)

* Refactor config * Update api keys * Create config * Update readme * Update device * Add require_api_key to check * Fix issue where saving configuration could result in file being cleared on failure * Roll Back tty_type Variable to model_type * fix: filter unsupported languages in get_lang_list * Update: Support for speaker parameter in Bert-VITS2 * Update: Changed input type of 'port' to text * Fix: Remove API key * Fix: Saving tts_models * Fix: speaker is null * Update: Add requests_toolbelt to requirements.txt * Update: bert-vits2 api test * Fix: null string * Fix: torch_data_type * Fix: espeak_library * Fix: stream_bert_vits2_infer lang error
Artrajz · Jan 2, 2025 · b16654f · b16654f
1 parent 58be4fd
commit b16654f
Show file tree

Hide file tree

Showing 49 changed files with 1,040 additions and 793 deletions.
diff --git a/README.md b/README.md
@@ -139,7 +139,7 @@ Starting from version 0.6.6, it is default behavior to automatically load all mo
 
 After the initial startup, a config.yaml configuration file will be generated. You need to change `tts_config.auto_load` to `false` in order to enable manual loading mode.
 
-You can modify the `tts_config.models` in the config.yaml or make modifications in the admin panel in the browser.
+You can modify the `tts_model_config.tts_modes` in the config.yaml or make modifications in the admin panel in the browser.
 
 **Note: After version 0.6.6, the model loading path has been modified. Please follow the steps below to configure the model path again!**
 
@@ -163,9 +163,9 @@ tts_config:
   auto_load: false
   models:
   - config_path: model1/config.json
-    model_path: model1/G_1000.pth
+    vits_path: model1/G_1000.pth
   - config_path: model2/config.json
-    model_path: model2/G_1000.pth
+    vits_path: model2/G_1000.pth
 	# GPT-SoVITS
   - sovits_path: gpt_sovits1/model1_e8_s11536.pth
     gpt_path: gpt_sovits1/model1-e15.ckpt
@@ -182,7 +182,7 @@ tts_config:
   auto_load: false
   models:
   - config_path: D://model3/config.json
-    model_path: D://model3/G_1000.pth
+    vits_path: D://model3/G_1000.pth
 ```
 
 - models_path: This is the models folder relative to the data directory, with the default value being "models". When auto_load is set to true, all models in the models_path directory will be loaded.
@@ -367,7 +367,7 @@ Please note that for the Chinese extra version, the version should be changed to
 
 ## API KEY
 
-Set `api_key_enabled: true` in `config.yaml` to enable API key authentication. The API key is `api_key: api-key`.
+Set `api_key_enabled: true` in `config.yaml` to enable API key authentication.
 After enabling it, you need to add the `api_key` parameter in GET requests and add the `X-API-KEY` parameter in the header for POST requests.
 
 # Parameter

diff --git a/README_zh.md b/README_zh.md
@@ -142,7 +142,7 @@ v0.6.6版本之后默认会**自动加载**`data/models`文件夹下的所有模
 
 首次启动之后会生成一个config.yaml配置文件，需要将`tts_config.auto_load`改为`false`以启用手动加载模式。
 
-可以修改配置文件中的`tts_config.models`或者在浏览器中进入管理员后台进行修改。
+可以修改配置文件中的`tts_model_config.tts_modes`或者在浏览器中进入管理员后台进行修改。
 
 **注意：v0.6.6版本之后已修改模型读取路径，请重新按照以下步骤配置模型路径！**
 
@@ -166,9 +166,9 @@ tts_config:
   auto_load: false
   models:
   - config_path: model1/config.json
-    model_path: model1/G_1000.pth
+    vits_path: model1/G_1000.pth
   - config_path: model2/config.json
-    model_path: model2/G_1000.pth
+    vits_path: model2/G_1000.pth
 	# GPT-SoVITS则为
   - sovits_path: gpt_sovits1/model1_e8_s11536.pth
     gpt_path: gpt_sovits1/model1-e15.ckpt
@@ -186,7 +186,7 @@ tts_config:
   auto_load: false
   models:
   - config_path: D://model3/config.json
-    model_path: D://model3/G_1000.pth
+    vits_path: D://model3/G_1000.pth
 ```
 
 - models_path:是相对于data目录下的模型文件夹，默认为models，auto_load为true时将会加载models_path目录下的所有模型。
@@ -379,7 +379,7 @@ url中的IP可在API启动后找到，一般使用192.168开头的局域网IP。
 
 ## API KEY
 
-在config.yaml中设置`api_key_enabled: true`以启用，api key填写：`api_key: api-key`。
+在config.yaml中设置`api_key_enabled: true`以启用。
 
 启用后，GET请求中使用需要增加参数api_key，POST请求中使用需要在header中添加参数`X-API-KEY`。
 

diff --git a/api_test.py b/api_test.py
@@ -349,11 +349,12 @@ def vits_json(text, id=0, format="wav", lang="auto", length=1, noise=0.667, nois
 
 
 # Bert_vits2
-def voice_bert_vits2(text, id=0, format="wav", lang="auto", length=1, noise=0.667, noisew=0.8, segment_size=50,
-                     sdp_ratio=0.2, save_audio=True, save_path=None):
+def voice_bert_vits2(text, id=0, speaker=None, format="wav", lang="auto", length=1, noise=0.667, noisew=0.8,
+                     segment_size=50, sdp_ratio=0.2, save_audio=True, save_path=None):
     fields = {
         "text": text,
         "id": str(id),
+        "speaker": str(speaker),
         "format": format,
         "lang": lang,
         "length": str(length),
@@ -568,8 +569,9 @@ def test_interface(text):
 
     # os.system(path)
 
-    # text = "你好“你的修炼速度有些出乎我的意料”"
+    # text = "你好“这是独处的妙处，我且受用这无边的荷香月色好了。”"
     # path = voice_reading_json(text=text, in_model_type="GPT-SOVITS", preset="wz", in_id=2, nr_model_type="BERT-VITS2",
     #                           nr_id=0)
 
+    # path = voice_bert_vits2("这是独处的妙处，我且受用这无边的荷香月色好了。", id=0, lang="zh", save_path=cache_path)
     # os.system(path)
diff --git a/app.py b/app.py
@@ -1,3 +1,4 @@
+import logger
 import logging
 import os.path
 
@@ -15,7 +16,7 @@
 from tts_app.auth.views import auth
 from tts_app.admin.views import admin
 
-from contants import config
+from config import config, BASE_DIR
 
 app = Flask(__name__, template_folder=os.path.join(os.path.dirname(__file__), 'tts_app', 'templates'),
             static_folder=os.path.join(os.path.dirname(__file__), 'tts_app', 'static'))
@@ -63,16 +64,16 @@ def create_folders(paths):
             os.makedirs(path, exist_ok=True)
 
 
-create_folders([os.path.join(config.abs_path, config.system.upload_folder),
-                os.path.join(config.abs_path, config.system.cache_path), ])
+create_folders([os.path.join(BASE_DIR, config.system.upload_folder),
+                os.path.join(BASE_DIR, config.system.cache_path), ])
 
 
 # regular cleaning
 @scheduler.task('interval', id='clean_task', seconds=config.system.clean_interval_seconds,
                 misfire_grace_time=900)
 def clean_task():
-    clean_folder(os.path.join(config.abs_path, config.system.upload_folder))
-    clean_folder(os.path.join(config.abs_path, config.system.cache_path))
+    clean_folder(os.path.join(BASE_DIR, config.system.upload_folder))
+    clean_folder(os.path.join(BASE_DIR, config.system.cache_path))
 
 
 if __name__ == '__main__':

diff --git a/bert_vits2/bert_vits2.py b/bert_vits2/bert_vits2.py
@@ -13,14 +13,14 @@
 from bert_vits2.text import *
 from bert_vits2.text.cleaner import clean_text
 from bert_vits2.utils import process_legacy_versions
-from contants import config
+from config import config
 from utils import get_hparams_from_file
 from utils.sentence import split_languages
 
 
 class Bert_VITS2:
-    def __init__(self, model_path, config, device=torch.device("cpu"), **kwargs):
-        self.model_path = model_path
+    def __init__(self, vits_path, config, device=torch.device("cpu"), **kwargs):
+        self.vits_path = vits_path
         self.hps_ms = get_hparams_from_file(config) if isinstance(config, str) else config
         self.n_speakers = getattr(self.hps_ms.data, 'n_speakers', 0)
         self.speakers = [item[0] for item in
@@ -220,7 +220,7 @@ def load_model(self, model_handler):
             zh_bert_extra=self.zh_bert_extra,
             **self.hps_ms.model).to(self.device)
         _ = self.net_g.eval()
-        bert_vits2_utils.load_checkpoint(self.model_path, self.net_g, None, skip_optimizer=True, version=self.version)
+        bert_vits2_utils.load_checkpoint(self.vits_path, self.net_g, None, skip_optimizer=True, version=self.version)
 
     def get_speakers(self):
         return self.speakers

diff --git a/bert_vits2/get_emo.py b/bert_vits2/get_emo.py
@@ -8,7 +8,7 @@
     Wav2Vec2PreTrainedModel,
 )
 
-from contants import config
+from config import config
 
 
 class RegressionHead(nn.Module):

diff --git a/bert_vits2/text/chinese_bert.py b/bert_vits2/text/chinese_bert.py
@@ -1,6 +1,6 @@
 import torch
 
-from contants import config
+from config import config
 
 
 def get_bert_feature(text, word2ph, tokenizer, model, device=config.system.device, style_text=None, style_weight=0.7,

diff --git a/bert_vits2/text/chinese_bert_extra.py b/bert_vits2/text/chinese_bert_extra.py
@@ -1,6 +1,6 @@
 import torch
 
-from contants import config
+from config import config
 
 
 def get_bert_feature(text, word2ph, tokenizer, model, device=config.system.device, style_text=None, style_weight=0.7,

diff --git a/bert_vits2/text/english_bert_mock.py b/bert_vits2/text/english_bert_mock.py
@@ -1,6 +1,6 @@
 import torch
 
-from contants import config
+from config import config
 
 
 def get_bert_feature(text, word2ph, tokenizer, model, device=config.system.device, style_text=None, style_weight=0.7,

diff --git a/bert_vits2/text/english_bert_mock_v200.py b/bert_vits2/text/english_bert_mock_v200.py
@@ -1,6 +1,6 @@
 import torch
 
-from contants import config
+from config import config
 
 
 def get_bert_feature(text, word2ph, tokenizer, model, device=config.system.device, **kwargs):

diff --git a/bert_vits2/text/japanese_bert.py b/bert_vits2/text/japanese_bert.py
@@ -1,6 +1,6 @@
 import torch
 
-from contants import config
+from config import config
 from bert_vits2.text.japanese import text2sep_kata
 
 LOCAL_PATH = "./bert/deberta-v2-large-japanese-char-wwm"

diff --git a/bert_vits2/text/japanese_bert_extra.py b/bert_vits2/text/japanese_bert_extra.py
@@ -1,6 +1,6 @@
 import torch
 
-from contants import config
+from config import config
 from bert_vits2.text.japanese import text2sep_kata
 
 

diff --git a/bert_vits2/text/japanese_bert_v111.py b/bert_vits2/text/japanese_bert_v111.py
@@ -1,6 +1,6 @@
 import torch
 
-from contants import config
+from config import config
 
 
 def get_bert_feature(text, word2ph, tokenizer, model, device=config.system.device, **kwargs):

diff --git a/bert_vits2/text/japanese_bert_v200.py b/bert_vits2/text/japanese_bert_v200.py
@@ -1,6 +1,6 @@
 import torch
 
-from contants import config
+from config import config
 from bert_vits2.text.japanese_v200 import text2sep_kata
 
 

diff --git a/bert_vits2/utils.py b/bert_vits2/utils.py
@@ -3,7 +3,7 @@
 import logging
 import torch
 
-from contants import config
+from config import config
 
 MATPLOTLIB_FLAG = False