Upload 17 files

app.py

@@ -362,7 +362,7 @@ if __name__ == "__main__":
     )
     speaker_ids = hps.data.spk2id
     speakers = list(speaker_ids.keys())
-    languages = ["ZH", "JP", "EN", "mix", "auto"]
+    languages = ["JP"]
     with gr.Blocks() as app:
         with gr.Row():
             with gr.Column():
@@ -370,13 +370,6 @@ if __name__ == "__main__":
                     label="输入文本内容",
                     placeholder="""
                     目前只支持日语！！
-                    如果你选择语言为\'mix\'，必须按照格式输入，否则报错:
-                        格式举例(zh是中文，jp是日语，不区分大小写；说话人举例:gongzi):
-                         [说话人1]<zh>你好，こんにちは！ <jp>こんにちは，世界。
-                         [说话人2]<zh>你好吗？<jp>元気ですか？
-                         [说话人3]<zh>谢谢。<jp>どういたしまして。
-                         ...
-                    另外，所有的语言选项都可以用'|'分割长段实现分句生成。
                     """,
                 )
                 slicer = gr.Button("快速切分", variant="primary")
@@ -399,7 +392,7 @@ if __name__ == "__main__":
                     minimum=0.1, maximum=2, value=1.0, step=0.1, label="Length"
                 )
                 language = gr.Dropdown(
-                    choices=languages, value=languages[1], label="Language"
+                    choices=languages, value=languages[0], label="Language"
                 )
                 btn = gr.Button("生成音频！", variant="primary")
             with gr.Column():

text/chinese_bert.py

@@ -5,8 +5,8 @@ from transformers import AutoModelForMaskedLM, AutoTokenizer
 
 from config import config
 
-#LOCAL_PATH = "./bert/chinese-roberta-wwm-ext-large"
-LOCAL_PATH = 'hfl/chinese-roberta-wwm-ext-large' 
+LOCAL_PATH = "./bert/chinese-roberta-wwm-ext-large"
+#LOCAL_PATH = 'hfl/chinese-roberta-wwm-ext-large' 
 
 tokenizer = AutoTokenizer.from_pretrained(LOCAL_PATH)
 

text/english.py

@@ -10,8 +10,8 @@ current_file_path = os.path.dirname(__file__)
 CMU_DICT_PATH = os.path.join(current_file_path, "cmudict.rep")
 CACHE_PATH = os.path.join(current_file_path, "cmudict_cache.pickle")
 _g2p = G2p()
-#LOCAL_PATH = "./bert/deberta-v3-large"
-LOCAL_PATH = 'microsoft/deberta-v3-large'
+LOCAL_PATH = "./bert/deberta-v3-large"
+#LOCAL_PATH = 'microsoft/deberta-v3-large'
 tokenizer = DebertaV2Tokenizer.from_pretrained(LOCAL_PATH)
 
 arpa = {

text/english_bert_mock.py

@@ -6,8 +6,8 @@ from transformers import DebertaV2Model, DebertaV2Tokenizer
 from config import config
 
 
-#LOCAL_PATH = "./bert/deberta-v3-large"
-LOCAL_PATH = 'microsoft/deberta-v2-xlarge'
+LOCAL_PATH = "./bert/deberta-v3-large"
+#LOCAL_PATH = 'microsoft/deberta-v2-xlarge'
 
 tokenizer = DebertaV2Tokenizer.from_pretrained(LOCAL_PATH)
 

text/japanese_bert.py

@@ -6,8 +6,8 @@ from transformers import AutoModelForMaskedLM, AutoTokenizer
 from config import config
 from text.japanese import text2sep_kata
 
-#LOCAL_PATH = "./bert/deberta-v2-large-japanese-char-wwm"
-LOCAL_PATH = 'ku-nlp/deberta-v2-large-japanese-char-wwm'
+LOCAL_PATH = "./bert/deberta-v2-large-japanese-char-wwm"
+#LOCAL_PATH = 'ku-nlp/deberta-v2-large-japanese-char-wwm'
 
 tokenizer = AutoTokenizer.from_pretrained(LOCAL_PATH)