update

app_locally.py

@@ -2,101 +2,25 @@ import os
 import torch
 import argparse
 import gradio as gr
-from zipfile import ZipFile
-import langid
-
-
-parser = argparse.ArgumentParser()
-parser.add_argument("--online_checkpoint_url", default="https://myshell-public-repo-hosting.s3.amazonaws.com/checkpoints_1226.zip")
-parser.add_argument("--share", action='store_true', default=False, help="make link public")
-args = parser.parse_args()
-
-# first download the checkpoints from server
-if not os.path.exists('checkpoints/'):
-    print('Downloading OpenVoice checkpoint ...')
-    os.system(f'wget {args.online_checkpoint_url} -O ckpt.zip')
-    print('Extracting OpenVoice checkpoint ...')
-    ZipFile("ckpt.zip").extractall()
+#from zipfile import ZipFile
+from melo.api import TTS
 
 # Init EN/ZH baseTTS and ToneConvertor
 from OpenVoice import se_extractor
 from OpenVoice.api import BaseSpeakerTTS, ToneColorConverter
+import devicetorch
 
-en_ckpt_base = 'checkpoints/base_speakers/EN'
-zh_ckpt_base = 'checkpoints/base_speakers/ZH'
-ckpt_converter = 'checkpoints/converter'
-device = 'cuda' if torch.cuda.is_available() else 'cpu'
-output_dir = 'outputs'
-os.makedirs(output_dir, exist_ok=True)
-en_base_speaker_tts = BaseSpeakerTTS(f'{en_ckpt_base}/config.json', device=device)
-en_base_speaker_tts.load_ckpt(f'{en_ckpt_base}/checkpoint.pth')
-zh_base_speaker_tts = BaseSpeakerTTS(f'{zh_ckpt_base}/config.json', device=device)
-zh_base_speaker_tts.load_ckpt(f'{zh_ckpt_base}/checkpoint.pth')
-tone_color_converter = ToneColorConverter(f'{ckpt_converter}/config.json', device=device)
-tone_color_converter.load_ckpt(f'{ckpt_converter}/checkpoint.pth')
-en_source_default_se = torch.load(f'{en_ckpt_base}/en_default_se.pth').to(device)
-en_source_style_se = torch.load(f'{en_ckpt_base}/en_style_se.pth').to(device)
-zh_source_se = torch.load(f'{zh_ckpt_base}/zh_default_se.pth').to(device)
+device = devicetorch.get(torch)
 
-supported_languages = ['zh', 'en']
-
-def predict(prompt, style, audio_file_pth, mic_file_path, use_mic, agree):
+def predict(prompt, style, audio_file_pth, mic_file_path, use_mic, language):
     # initialize a empty info
     text_hint = ''
-    # agree with the terms
-    if agree == False:
-        text_hint += '[ERROR] Please accept the Terms & Condition!\n'
-        gr.Warning("Please accept the Terms & Condition!")
-        return (
-            text_hint,
-            None,
-            None,
-        )
 
-    # first detect the input language
-    language_predicted = langid.classify(prompt)[0].strip()  
-    print(f"Detected language:{language_predicted}")
-
-    if language_predicted not in supported_languages:
-        text_hint += f"[ERROR] The detected language {language_predicted} for your input text is not in our Supported Languages: {supported_languages}\n"
-        gr.Warning(
-            f"The detected language {language_predicted} for your input text is not in our Supported Languages: {supported_languages}"
-        )
+    tts_model = TTS(language=language, device=device)
+    speaker_id = models[language].hps.data.spk2id
+    speaker_key = speaker_key.lower().replace('_', '-')
+    source_se = torch.load(f'checkpoints/base_speakers/ses/{speaker_key}.pth', map_location=device)
 
-        return (
-            text_hint,
-            None,
-            None,
-        )
-    
-    if language_predicted == "zh":
-        tts_model = zh_base_speaker_tts
-        source_se = zh_source_se
-        language = 'Chinese'
-        if style not in ['default']:
-            text_hint += f"[ERROR] The style {style} is not supported for Chinese, which should be in ['default']\n"
-            gr.Warning(f"The style {style} is not supported for Chinese, which should be in ['default']")
-            return (
-                text_hint,
-                None,
-                None,
-            )
-
-    else:
-        tts_model = en_base_speaker_tts
-        if style == 'default':
-            source_se = en_source_default_se
-        else:
-            source_se = en_source_style_se
-        language = 'English'
-        if style not in ['default', 'whispering', 'shouting', 'excited', 'cheerful', 'terrified', 'angry', 'sad', 'friendly']:
-            text_hint += f"[ERROR] The style {style} is not supported for English, which should be in ['default', 'whispering', 'shouting', 'excited', 'cheerful', 'terrified', 'angry', 'sad', 'friendly']\n"
-            gr.Warning(f"The style {style} is not supported for English, which should be in ['default', 'whispering', 'shouting', 'excited', 'cheerful', 'terrified', 'angry', 'sad', 'friendly']")
-            return (
-                text_hint,
-                None,
-                None,
-            )
 
     if use_mic == True:
         if mic_file_path is not None:
@@ -123,16 +47,6 @@ def predict(prompt, style, audio_file_pth, mic_file_path, use_mic, agree):
             None,
             None,
         )
-#    if len(prompt) > 200:
-#        text_hint += f"[ERROR] Text length limited to 200 characters for this demo, please try shorter text. You can clone our open-source repo and try for your usage \n"
-#        gr.Warning(
-#            "Text length limited to 200 characters for this demo, please try shorter text. You can clone our open-source repo for your usage"
-#        )
-#        return (
-#            text_hint,
-#            None,
-#            None,
-#        )
     
     # note diffusion_conditioning not used on hifigan (default mode), it will be empty but need to pass it to model.inference
     try:
@@ -150,6 +64,9 @@ def predict(prompt, style, audio_file_pth, mic_file_path, use_mic, agree):
         )
 
     src_path = f'{output_dir}/tmp.wav'
+
+    speed = 1.0
+
     tts_model.tts(prompt, src_path, speaker=style, language=language)
 
     save_path = f'{output_dir}/output.wav'
@@ -171,46 +88,6 @@ def predict(prompt, style, audio_file_pth, mic_file_path, use_mic, agree):
     )
 
 
-
-title = "MyShell OpenVoice"
-
-description = """
-We introduce OpenVoice, a versatile instant voice cloning approach that requires only a short audio clip from the reference speaker to replicate their voice and generate speech in multiple languages. OpenVoice enables granular control over voice styles, including emotion, accent, rhythm, pauses, and intonation, in addition to replicating the tone color of the reference speaker. OpenVoice also achieves zero-shot cross-lingual voice cloning for languages not included in the massive-speaker training set.
-"""
-
-markdown_table = """
-<div align="center" style="margin-bottom: 10px;">
-
-|               |               |               |
-| :-----------: | :-----------: | :-----------: | 
-| **OpenSource Repo** | **Project Page** | **Join the Community** |        
-| <div style='text-align: center;'><a style="display:inline-block,align:center" href='https://github.com/myshell-ai/OpenVoice'><img src='https://img.shields.io/github/stars/myshell-ai/OpenVoice?style=social' /></a></div> | [OpenVoice](https://research.myshell.ai/open-voice) | [![Discord](https://img.shields.io/discord/1122227993805336617?color=%239B59B6&label=%20Discord%20)](https://discord.gg/myshell) |
-
-</div>
-"""
-
-markdown_table_v2 = """
-<div align="center" style="margin-bottom: 2px;">
-
-|               |               |               |              |
-| :-----------: | :-----------: | :-----------: | :-----------: | 
-| **OpenSource Repo** | <div style='text-align: center;'><a style="display:inline-block,align:center" href='https://github.com/myshell-ai/OpenVoice'><img src='https://img.shields.io/github/stars/myshell-ai/OpenVoice?style=social' /></a></div> |  **Project Page** |  [OpenVoice](https://research.myshell.ai/open-voice) |     
-
-| | |
-| :-----------: | :-----------: |
-**Join the Community** |   [![Discord](https://img.shields.io/discord/1122227993805336617?color=%239B59B6&label=%20Discord%20)](https://discord.gg/myshell) |
-
-</div>
-"""
-content = """
-<div>
-  <strong>For multi-lingual & cross-lingual examples, please refer to <a href='https://github.com/myshell-ai/OpenVoice/blob/main/demo_part2.ipynb'>this jupyter notebook</a>.</strong>
-  This online demo mainly supports <strong>English</strong>. The <em>default</em> style also supports <strong>Chinese</strong>. But OpenVoice can adapt to any other language as long as a base speaker is provided.
-</div>
-"""
-wrapped_markdown_content = f"<div style='border: 1px solid #000; padding: 10px;'>{content}</div>"
-
-
 examples = [
     [
         "今天天气真好，我们一起出去吃饭吧。",
@@ -239,8 +116,8 @@ examples = [
 
 with gr.Blocks(analytics_enabled=False) as demo:
 
-    with gr.Row():
-        gr.HTML(wrapped_markdown_content)
+#    with gr.Row():
+#        gr.HTML(wrapped_markdown_content)
 
     with gr.Row():
         with gr.Column():
@@ -273,11 +150,7 @@ with gr.Blocks(analytics_enabled=False) as demo:
                 value=False,
                 info="Notice: Microphone input may not work properly under traffic",
             )
-            tos_gr = gr.Checkbox(
-                label="Agree",
-                value=False,
-                info="I agree to the terms of the cc-by-nc-4.0 license-: https://github.com/myshell-ai/OpenVoice/blob/main/LICENSE",
-            )
+            language = gr.Radio(['EN_NEWEST', 'EN', 'ES', 'FR', 'ZH', 'JP', 'KR'], label='Language', value='EN_NEWEST')
 
             tts_button = gr.Button("Send", elem_id="send-btn", visible=True)
 
@@ -289,11 +162,11 @@ with gr.Blocks(analytics_enabled=False) as demo:
 
             gr.Examples(examples,
                         label="Examples",
-                        inputs=[input_text_gr, style_gr, ref_gr, mic_gr, use_mic_gr, tos_gr],
+                        inputs=[input_text_gr, style_gr, ref_gr, mic_gr, use_mic_gr, language],
                         outputs=[out_text_gr, audio_gr, ref_audio_gr],
                         fn=predict,
                         cache_examples=False,)
-            tts_button.click(predict, [input_text_gr, style_gr, ref_gr, mic_gr, use_mic_gr, tos_gr], outputs=[out_text_gr, audio_gr, ref_audio_gr])
+            tts_button.click(predict, [input_text_gr, style_gr, ref_gr, mic_gr, use_mic_gr, language], outputs=[out_text_gr, audio_gr, ref_audio_gr])
 
 demo.queue()  
-demo.launch(debug=True, show_api=True, share=args.share)
+demo.launch(debug=True, show_api=True)

requirements_locally.txt

@@ -1,18 +1,18 @@
-langid
-librosa==0.9.1
-faster-whisper
-pydub==0.25.1
-#wavmark==0.0.2
-wavmark
-#numpy==1.22.0
-numpy
-eng_to_ipa==0.0.2
-inflect==7.0.0
-unidecode==1.3.7
-whisper-timestamped
-openai
-python-dotenv
-pypinyin==0.50.0
-cn2an==0.5.22
-jieba==0.42.1
+#langid
+#librosa==0.9.1
+#faster-whisper
+#pydub==0.25.1
+##wavmark==0.0.2
+#wavmark
+##numpy==1.22.0
+#numpy
+#eng_to_ipa==0.0.2
+#inflect==7.0.0
+#unidecode==1.3.7
+#whisper-timestamped
+#openai
+#python-dotenv
+#pypinyin==0.50.0
+#cn2an==0.5.22
+#jieba==0.42.1
 gradio==3.48.0