Spaces:
Running
Running
| import gradio as gr | |
| from rvc import * | |
| def change_audio_mode(vc_audio_mode): | |
| if vc_audio_mode == "Input path": | |
| return ( | |
| # Input & Upload | |
| gr.Textbox.update(visible=True), | |
| gr.Audio.update(visible=False), | |
| # Youtube | |
| gr.Dropdown.update(visible=False), | |
| gr.Textbox.update(visible=False), | |
| gr.Dropdown.update(visible=False), | |
| gr.Button.update(visible=False), | |
| gr.Audio.update(visible=False), | |
| gr.Audio.update(visible=False), | |
| gr.Audio.update(visible=False), | |
| gr.Slider.update(visible=False), | |
| gr.Audio.update(visible=False), | |
| gr.Button.update(visible=False), | |
| # TTS | |
| gr.Textbox.update(visible=False), | |
| gr.Dropdown.update(visible=False) | |
| ) | |
| elif vc_audio_mode == "Upload audio": | |
| return ( | |
| # Input & Upload | |
| gr.Textbox.update(visible=False), | |
| gr.Audio.update(visible=True), | |
| # Youtube | |
| gr.Dropdown.update(visible=False), | |
| gr.Textbox.update(visible=False), | |
| gr.Dropdown.update(visible=False), | |
| gr.Button.update(visible=False), | |
| gr.Audio.update(visible=False), | |
| gr.Audio.update(visible=False), | |
| gr.Audio.update(visible=False), | |
| gr.Slider.update(visible=False), | |
| gr.Audio.update(visible=False), | |
| gr.Button.update(visible=False), | |
| # TTS | |
| gr.Textbox.update(visible=False), | |
| gr.Dropdown.update(visible=False) | |
| ) | |
| elif vc_audio_mode == "Youtube": | |
| return ( | |
| # Input & Upload | |
| gr.Textbox.update(visible=False), | |
| gr.Audio.update(visible=False), | |
| # Youtube | |
| gr.Dropdown.update(visible=True), | |
| gr.Textbox.update(visible=True), | |
| gr.Dropdown.update(visible=True), | |
| gr.Button.update(visible=True), | |
| gr.Audio.update(visible=True), | |
| gr.Audio.update(visible=True), | |
| gr.Audio.update(visible=True), | |
| gr.Slider.update(visible=True), | |
| gr.Audio.update(visible=True), | |
| gr.Button.update(visible=True), | |
| # TTS | |
| gr.Textbox.update(visible=False), | |
| gr.Dropdown.update(visible=False) | |
| ) | |
| elif vc_audio_mode == "TTS Audio": | |
| return ( | |
| # Input & Upload | |
| gr.Textbox.update(visible=False), | |
| gr.Audio.update(visible=False), | |
| # Youtube | |
| gr.Dropdown.update(visible=False), | |
| gr.Textbox.update(visible=False), | |
| gr.Dropdown.update(visible=False), | |
| gr.Button.update(visible=False), | |
| gr.Audio.update(visible=False), | |
| gr.Audio.update(visible=False), | |
| gr.Audio.update(visible=False), | |
| gr.Slider.update(visible=False), | |
| gr.Audio.update(visible=False), | |
| gr.Button.update(visible=False), | |
| # TTS | |
| gr.Textbox.update(visible=True), | |
| gr.Dropdown.update(visible=True) | |
| ) | |
| else: | |
| return ( | |
| # Input & Upload | |
| gr.Textbox.update(visible=False), | |
| gr.Audio.update(visible=True), | |
| # Youtube | |
| gr.Dropdown.update(visible=False), | |
| gr.Textbox.update(visible=False), | |
| gr.Dropdown.update(visible=False), | |
| gr.Button.update(visible=False), | |
| gr.Audio.update(visible=False), | |
| gr.Audio.update(visible=False), | |
| gr.Audio.update(visible=False), | |
| gr.Slider.update(visible=False), | |
| gr.Audio.update(visible=False), | |
| gr.Button.update(visible=False), | |
| # TTS | |
| gr.Textbox.update(visible=False), | |
| gr.Dropdown.update(visible=False) | |
| ) | |
| if __name__ == '__main__': | |
| load_hubert() | |
| categories = load_model() | |
| tts_voice_list = asyncio.get_event_loop().run_until_complete(edge_tts.list_voices()) | |
| voices = [f"{v['ShortName']}-{v['Gender']}" for v in tts_voice_list] | |
| with gr.Blocks(theme="nevreal/blues") as app: | |
| gr.Markdown( | |
| "# <center> Hololive RVC Models\n" | |
| "[](https://colab.research.google.com/github/aziib/hololive-rvc-models-v2/blob/main/hololive_rvc_models_v2.ipynb)\n\n" | |
| ) | |
| for (folder_title, folder, models) in categories: | |
| with gr.TabItem(folder_title): | |
| with gr.Tabs(): | |
| if not models: | |
| gr.Markdown("# <center> No Model Loaded.") | |
| gr.Markdown("## <center> Please add model or fix your model path.") | |
| continue | |
| for (name, title, author, cover, model_version, vc_fn) in models: | |
| with gr.TabItem(name): | |
| with gr.Row(): | |
| gr.Markdown( | |
| '<div align="center">' | |
| f'<div>{title}</div>\n'+ | |
| f'<div>RVC {model_version} Model</div>\n'+ | |
| #(f'<div>Model author: {author}</div>' if author else "")+ | |
| #(f'<img style="width:auto;height:300px;" src="file/{cover}">' if cover else "")+ | |
| '</div>' | |
| ) | |
| with gr.Row(): | |
| with gr.Column(): | |
| with gr.Accordion("Main Options", open=False): | |
| vc_audio_mode = gr.Dropdown(label="Input voice", choices=audio_mode, allow_custom_value=False, value="Upload audio") | |
| # Input and Upload | |
| vc_input = gr.Textbox(label="Input audio path", visible=False) | |
| vc_upload = gr.Audio(label="Upload audio file", visible=True, interactive=True) | |
| # Youtube | |
| vc_download_audio = gr.Dropdown(label="Provider", choices=["Youtube"], allow_custom_value=False, visible=False, value="Youtube", info="Select provider (Default: Youtube)") | |
| vc_link = gr.Textbox(label="Youtube URL", visible=False, info="Example: https://www.youtube.com/watch?v=Nc0sB1Bmf-A", placeholder="https://www.youtube.com/watch?v=...") | |
| vc_split_model = gr.Dropdown(label="Splitter Model", choices=["htdemucs", "mdx_extra_q"], allow_custom_value=False, visible=False, value="htdemucs", info="Select the splitter model (Default: htdemucs)") | |
| vc_split = gr.Button("Split Audio", variant="primary", visible=False) | |
| vc_vocal_preview = gr.Audio(label="Vocal Preview", visible=False) | |
| vc_inst_preview = gr.Audio(label="Instrumental Preview", visible=False) | |
| vc_audio_preview = gr.Audio(label="Audio Preview", visible=False) | |
| # TTS | |
| tts_text = gr.Textbox(visible=False, label="TTS text", info="Text to speech input") | |
| tts_voice = gr.Dropdown(label="Edge-tts speaker", choices=voices, visible=False, allow_custom_value=False, value="en-US-AnaNeural-Female") | |
| with gr.Column(): | |
| vc_transform0 = gr.Number(label="Transpose", value=0, info='Type "12" to change from male to female voice. Type "-12" to change female to male voice') | |
| f0method0 = gr.Radio( | |
| label="Pitch extraction algorithm", | |
| info=f0method_info, | |
| choices=f0method_mode, | |
| value="rmvpe", | |
| interactive=True | |
| ) | |
| index_rate1 = gr.Slider( | |
| minimum=0, | |
| maximum=1, | |
| label="Retrieval feature ratio", | |
| info="Accents controling. Too high prob gonna sounds too robotic (Default: 0.4)", | |
| value=0.4, | |
| interactive=True, | |
| ) | |
| filter_radius0 = gr.Slider( | |
| minimum=0, | |
| maximum=7, | |
| label="Apply Median Filtering", | |
| info="The value represents the filter radius and can reduce breathiness.", | |
| value=1, | |
| step=1, | |
| interactive=True, | |
| ) | |
| resample_sr0 = gr.Slider( | |
| minimum=0, | |
| maximum=48000, | |
| label="Resample the output audio", | |
| info="Resample the output audio in post-processing to the final sample rate. Set to 0 for no resampling", | |
| value=0, | |
| step=1, | |
| interactive=True, | |
| ) | |
| rms_mix_rate0 = gr.Slider( | |
| minimum=0, | |
| maximum=1, | |
| label="Volume Envelope", | |
| info="Use the volume envelope of the input to replace or mix with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is used", | |
| value=1, | |
| interactive=True, | |
| ) | |
| protect0 = gr.Slider( | |
| minimum=0, | |
| maximum=0.5, | |
| label="Voice Protection", | |
| info="Protect voiceless consonants and breath sounds to prevent artifacts such as tearing in electronic music. Set to 0.5 to disable. Decrease the value to increase protection, but it may reduce indexing accuracy", | |
| value=0.23, | |
| step=0.01, | |
| interactive=True, | |
| ) | |
| with gr.Column(): | |
| vc_log = gr.Textbox(label="Output Information", interactive=False) | |
| vc_output = gr.Audio(label="Output Audio", interactive=False) | |
| vc_convert = gr.Button("Convert", variant="primary") | |
| vc_volume = gr.Slider( | |
| minimum=0, | |
| maximum=10, | |
| label="Vocal volume", | |
| value=4, | |
| interactive=True, | |
| step=1, | |
| info="Adjust vocal volume (Default: 4}", | |
| visible=False | |
| ) | |
| vc_combined_output = gr.Audio(label="Output Combined Audio", visible=False) | |
| vc_combine = gr.Button("Combine",variant="primary", visible=False) | |
| vc_convert.click( | |
| fn=vc_fn, | |
| inputs=[ | |
| vc_audio_mode, | |
| vc_input, | |
| vc_upload, | |
| tts_text, | |
| tts_voice, | |
| vc_transform0, | |
| f0method0, | |
| index_rate1, | |
| filter_radius0, | |
| resample_sr0, | |
| rms_mix_rate0, | |
| protect0, | |
| ], | |
| outputs=[vc_log ,vc_output] | |
| ) | |
| vc_split.click( | |
| fn=cut_vocal_and_inst, | |
| inputs=[vc_link, vc_download_audio, vc_split_model], | |
| outputs=[vc_vocal_preview, vc_inst_preview, vc_audio_preview, vc_input] | |
| ) | |
| vc_combine.click( | |
| fn=combine_vocal_and_inst, | |
| inputs=[vc_output, vc_volume, vc_split_model], | |
| outputs=[vc_combined_output] | |
| ) | |
| vc_audio_mode.change( | |
| fn=change_audio_mode, | |
| inputs=[vc_audio_mode], | |
| outputs=[ | |
| vc_input, | |
| vc_upload, | |
| vc_download_audio, | |
| vc_link, | |
| vc_split_model, | |
| vc_split, | |
| vc_vocal_preview, | |
| vc_inst_preview, | |
| vc_audio_preview, | |
| vc_volume, | |
| vc_combined_output, | |
| vc_combine, | |
| tts_text, | |
| tts_voice | |
| ] | |
| ) | |
| if limitation is True: | |
| app.queue(concurrency_count=1, max_size=20, api_open=config.api).launch(share=config.colab) | |
| else: | |
| app.queue(concurrency_count=1, max_size=20, api_open=config.api).launch(share=True) |