Update app.py

app.py

@@ -46,7 +46,8 @@ logging.getLogger("infer_rvc_python").setLevel(logging.ERROR)
 converter = BaseLoader(only_cpu=False, hubert_path=None, rmvpe_path=None)
 converter.hu_bert_model = load_hu_bert(Config(only_cpu=False), converter.hubert_path)
 
-test_model = "https://huggingface.co/sail-rvc/Aldeano_Minecraft__RVC_V2_-_500_Epochs_/resolve/main/model.pth?download=true, https://huggingface.co/sail-rvc/Aldeano_Minecraft__RVC_V2_-_500_Epochs_/resolve/main/model.index?download=true"
+# مدل پیش‌فرض (اختیاری)
+test_model = "https://huggingface.co/sail-rvc/Aldeano_Minecraft__RVC_V2_-_500_Epochs_/resolve/main/model.pth?download=true,  https://huggingface.co/sail-rvc/Aldeano_Minecraft__RVC_V2_-_500_Epochs_/resolve/main/model.index?download=true"
 test_names = ["model.pth", "model.index"]
 
 for url, filename in zip(test_model.split(", "), test_names):
@@ -64,9 +65,9 @@ for url, filename in zip(test_model.split(", "), test_names):
         with open(filename, "wb") as f:
             pass
 
-title = "<center><strong><font size='7'>RVC⚡ZERO</font></strong></center>"
-description = "This demo is provided for educational and research purposes only. The authors and contributors of this project do not endorse or encourage any misuse or unethical use of this software. Any use of this software for purposes other than those intended is solely at the user's own risk. The authors and contributors shall not be held responsible for any damages or liabilities arising from the use of this demo inappropriately." if IS_ZERO_GPU else ""
-RESOURCES = "- You can also try `RVC⚡ZERO` in Colab’s free tier, which provides free GPU [link](https://github.com/R3gm/rvc_zero_ui?tab=readme-ov-file#rvczero)."
+title = "<center><strong><font size='7'>RVC⚡ZERO - Local Upload</font></strong></center>"
+description = "Upload your own model (.pth) and audio files for voice conversion." if IS_ZERO_GPU else ""
+RESOURCES = "- You can upload your custom RVC models and audio files directly."
 theme = args.theme
 delete_cache_time = (3200, 3200) if IS_ZERO_GPU else (86400, 86400)
 
@@ -83,15 +84,15 @@ async def get_voices_list(proxy=None):
     """Print all available voices."""
     from edge_tts import list_voices
     voices = await list_voices(proxy=proxy)
-    voices = sorted(voices, key=lambda voice: voice["ShortName"])
+    voices = sorted(voices, key=lambda voice: voice.get("ShortName", ""))
 
     table = [
         {
-            "ShortName": voice["ShortName"],
-            "Gender": voice["Gender"],
-            "ContentCategories": ", ".join(voice["VoiceTag"]["ContentCategories"]),
-            "VoicePersonalities": ", ".join(voice["VoiceTag"]["VoicePersonalities"]),
-            "FriendlyName": voice["FriendlyName"],
+            "ShortName": voice.get("ShortName", "Unknown"),
+            "Gender": voice.get("Gender", "Unknown"),
+            "ContentCategories": ", ".join(voice.get("VoiceTag", {}).get("ContentCategories", [])),
+            "VoicePersonalities": ", ".join(voice.get("VoiceTag", {}).get("VoicePersonalities", [])),
+            "FriendlyName": voice.get("FriendlyName", voice.get("Name", "Unknown Voice")),
         }
         for voice in voices
     ]
@@ -102,11 +103,8 @@ async def get_voices_list(proxy=None):
 def find_files(directory):
     file_paths = []
     for filename in os.listdir(directory):
-        # Check if the file has the desired extension
         if filename.endswith('.pth') or filename.endswith('.zip') or filename.endswith('.index'):
-            # If yes, add the file path to the list
             file_paths.append(os.path.join(directory, filename))
-
     return file_paths
 
 
@@ -120,7 +118,6 @@ def unzip_in_folder(my_zip, my_dir):
 
 
 def find_my_model(a_, b_):
-
     if a_ is None or a_.endswith(".pth"):
         return a_, b_
 
@@ -179,7 +176,6 @@ def ensure_valid_file(url):
             raise ValueError("No Content-Length header found")
 
         file_size = int(content_length)
-        # print("debug", url, file_size)
         if file_size > 900000000 and IS_ZERO_GPU:
             raise ValueError("The file is too large. Max allowed is 900 MB.")
 
@@ -196,11 +192,10 @@ def clear_files(directory):
 
 
 def get_my_model(url_data, progress=gr.Progress(track_tqdm=True)):
-
-    if not url_data:
+    if not url_
         return None, None
 
-    if "," in url_data:
+    if "," in url_
         a_, b_ = url_data.split(",")
         a_, b_ = a_.strip().replace("/blob/", "/resolve/"), b_.strip().replace("/blob/", "/resolve/")
     else:
@@ -250,8 +245,6 @@ def get_my_model(url_data, progress=gr.Progress(track_tqdm=True)):
     except Exception as e:
         raise e
     finally:
-        # time.sleep(10)
-        # shutil.rmtree(directory)
         t = threading.Thread(target=clear_files, args=(directory,))
         t.start()
 
@@ -264,7 +257,6 @@ def add_audio_effects(audio_list, type_output):
         try:
             output_path = f'{os.path.splitext(audio_path)[0]}_effects.{type_output}'
 
-            # Initialize audio effects plugins
             board = Pedalboard(
                 [
                     HighpassFilter(),
@@ -273,7 +265,6 @@ def add_audio_effects(audio_list, type_output):
                  ]
             )
 
-            # Temporary WAV to hold processed data before exporting
             temp_wav = f'{os.path.splitext(audio_path)[0]}_temp.wav'
 
             with AudioFile(audio_path) as f:
@@ -283,11 +274,9 @@ def add_audio_effects(audio_list, type_output):
                         effected = board(chunk, f.samplerate, reset=False)
                         o.write(effected)
 
-            # Convert with pydub to desired output type
             audio_seg = AudioSegment.from_file(temp_wav, format=type_output)
             audio_seg.export(output_path, format=type_output, bitrate=("320k" if type_output == "mp3" else None))
 
-            # Clean up temp file
             os.remove(temp_wav)
 
             result.append(output_path)
@@ -300,7 +289,6 @@ def add_audio_effects(audio_list, type_output):
 
 
 def apply_noisereduce(audio_list, type_output):
-    # https://github.com/sa-if/Audio-Denoiser
     print("Noice reduce")
 
     result = []
@@ -308,16 +296,10 @@ def apply_noisereduce(audio_list, type_output):
         out_path = f"{os.path.splitext(audio_path)[0]}_noisereduce.{type_output}"
 
         try:
-            # Load audio file
             audio = AudioSegment.from_file(audio_path)
-
-            # Convert audio to numpy array
             samples = np.array(audio.get_array_of_samples())
-
-            # Reduce noise
             reduced_noise = nr.reduce_noise(samples, sr=audio.frame_rate, prop_decrease=0.6)
 
-            # Convert reduced noise signal back to audio
             reduced_audio = AudioSegment(
                 reduced_noise.tobytes(), 
                 frame_rate=audio.frame_rate, 
@@ -325,7 +307,6 @@ def apply_noisereduce(audio_list, type_output):
                 channels=audio.channels
             )
 
-            # Save reduced audio to file
             reduced_audio.export(out_path, format=type_output, bitrate=("320k" if type_output == "mp3" else None))
             result.append(out_path)
 
@@ -367,7 +348,7 @@ def run(
     steps,
 ):
     if not audio_files:
-        raise ValueError("The audio pls")
+        raise ValueError("Please upload audio files")
 
     if isinstance(audio_files, str):
         audio_files = [audio_files]
@@ -411,17 +392,19 @@ def run(
 
 def audio_conf():
     return gr.File(
-        label="Audio files",
+        label="Upload Audio Files (wav, mp3, ogg, flac)",
         file_count="multiple",
         type="filepath",
+        file_types=[".wav", ".mp3", ".ogg", ".flac", ".m4a"],
         container=True,
     )
 
 
 def model_conf():
     return gr.File(
-        label="Model file",
+        label="Upload Model File (.pth)",
         type="filepath",
+        file_types=[".pth"],
         height=130,
     )
 
@@ -450,8 +433,9 @@ def pitch_lvl_conf():
 
 def index_conf():
     return gr.File(
-        label="Index file",
+        label="Upload Index File (.index) - Optional",
         type="filepath",
+        file_types=[".index"],
         height=130,
     )
 
@@ -498,14 +482,15 @@ def consonant_protec_conf():
 
 def button_conf():
     return gr.Button(
-        "Inference",
+        "Convert Voice",
         variant="primary",
+        size="lg",
     )
 
 
 def output_conf():
     return gr.File(
-        label="Result",
+        label="Converted Audio",
         file_count="multiple",
         interactive=False,
     )
@@ -514,25 +499,24 @@ def output_conf():
 def active_tts_conf():
     return gr.Checkbox(
         False,
-        label="TTS",
-        # info="",
+        label="Use Text-to-Speech",
         container=False,
     )
 
 
 def tts_voice_conf():
     return gr.Dropdown(
-        label="tts voice",
-        choices=voices,
+        label="TTS Voice",
+        choices=[],  # Will be populated later
         visible=False,
-        value="en-US-EmmaMultilingualNeural-Female",
+        value=None,
     )
 
 
 def tts_text_conf():
     return gr.Textbox(
         value="",
-        placeholder="Write the text here...",
+        placeholder="Enter text to convert to speech...",
         label="Text",
         visible=False,
         lines=3,
@@ -541,7 +525,7 @@ def tts_text_conf():
 
 def tts_button_conf():
     return gr.Button(
-        "Process TTS",
+        "Generate Speech",
         variant="secondary",
         visible=False,
     )
@@ -550,8 +534,7 @@ def tts_button_conf():
 def tts_play_conf():
     return gr.Checkbox(
         False,
-        label="Play",
-        # info="",
+        label="Auto-play generated audio",
         container=False,
         visible=False,
     )
@@ -561,7 +544,6 @@ def sound_gui():
     return gr.Audio(
         value=None,
         type="filepath",
-        # format="mp3",
         autoplay=True,
         visible=True,
         interactive=False,
@@ -582,16 +564,16 @@ def steps_conf():
 
 def format_output_gui():
     return gr.Dropdown(
-        label="Format output:",
+        label="Output Format:",
         choices=["wav", "mp3", "flac"],
         value="wav",
     )
 
+
 def denoise_conf():
     return gr.Checkbox(
         False,
-        label="Denoise",
-        # info="",
+        label="Apply Noise Reduction",
         container=False,
         visible=True,
     )
@@ -600,8 +582,7 @@ def denoise_conf():
 def effects_conf():
     return gr.Checkbox(
         False,
-        label="Reverb",
-        # info="",
+        label="Apply Audio Effects (Reverb)",
         container=False,
         visible=True,
     )
@@ -615,7 +596,13 @@ def infer_tts_audio(tts_voice, tts_text, play_tts):
     os.makedirs(os.path.join(out_dir, folder_tts), exist_ok=True)
     out_path = os.path.join(out_dir, folder_tts, "tts.mp3")
 
-    asyncio.run(edge_tts.Communicate(tts_text, "-".join(tts_voice.split('-')[:-1])).save(out_path))
+    # Extract ShortName from combined value (e.g., "en-US-EmmaMultilingualNeural-Female")
+    if tts_voice:
+        short_name = "-".join(tts_voice.split('-')[:-1])
+    else:
+        short_name = "en-US-EmmaMultilingualNeural"
+
+    asyncio.run(edge_tts.Communicate(tts_text, short_name).save(out_path))
     if play_tts:
         return [out_path], out_path
     return [out_path], None
@@ -636,8 +623,7 @@ def show_components_tts(value_active):
 def down_active_conf():
     return gr.Checkbox(
         False,
-        label="URL-to-Model",
-        # info="",
+        label="Download from URL",
         container=False,
     )
 
@@ -645,8 +631,8 @@ def down_active_conf():
 def down_url_conf():
     return gr.Textbox(
         value="",
-        placeholder="Write the url here...",
-        label="Enter URL",
+        placeholder="Hugging Face model URL...",
+        label="Model URL",
         visible=False,
         lines=1,
     )
@@ -654,7 +640,7 @@ def down_url_conf():
 
 def down_button_conf():
     return gr.Button(
-        "Process",
+        "Download Model",
         variant="secondary",
         visible=False,
     )
@@ -671,7 +657,7 @@ def show_components_down(value_active):
 
 CSS = """
 #audio_tts {
-  visibility: hidden;   /* invisible but still takes space */
+  visibility: hidden;
   height: 0px;
   width: 0px;
   max-width: 0px;
@@ -684,168 +670,146 @@ def get_gui(theme):
         gr.Markdown(title)
         gr.Markdown(description)
 
-        active_tts = active_tts_conf()
-        with gr.Row():
-            with gr.Column(scale=1):
-                tts_text = tts_text_conf()
-            with gr.Column(scale=2):
-                with gr.Row():
-                    with gr.Column():
-                        with gr.Row():
-                            tts_voice = tts_voice_conf()
-                            tts_active_play = tts_play_conf()
-
-                tts_button = tts_button_conf()
-                tts_play = sound_gui()
-
-        active_tts.change(
-            fn=show_components_tts,
-            inputs=[active_tts],
-            outputs=[tts_voice, tts_text, tts_button, tts_active_play],
-        )
-
-        aud = audio_conf()
-        # gr.HTML("<hr>")
-
-        tts_button.click(
-            fn=infer_tts_audio,
-            inputs=[tts_voice, tts_text, tts_active_play],
-            outputs=[aud, tts_play],
-        )
+        with gr.Tab("Voice Conversion"):
+            # بخش آپلود فایل‌های صوتی
+            gr.Markdown("### 📤 Upload Audio Files")
+            aud = audio_conf()
+            
+            # بخش TTS
+            active_tts = active_tts_conf()
+            with gr.Row(visible=False) as tts_row:
+                with gr.Column(scale=1):
+                    tts_text = tts_text_conf()
+                with gr.Column(scale=2):
+                    with gr.Row():
+                        with gr.Column():
+                            with gr.Row():
+                                tts_voice = tts_voice_conf()
+                                tts_active_play = tts_play_conf()
+                    tts_button = tts_button_conf()
+                    tts_play = sound_gui()
+
+            active_tts.change(
+                fn=show_components_tts,
+                inputs=[active_tts],
+                outputs=[tts_voice, tts_text, tts_button, tts_active_play],
+            )
 
-        down_active_gui = down_active_conf()
-        down_info = gr.Markdown(
-            f"Provide a link to a zip file, like this one: `https://huggingface.co/MrDawg/ToothBrushing/resolve/main/ToothBrushing.zip?download=true`, or separate links with a comma for the .pth and .index files, like this: `{test_model}`",
-            visible=False
-        )
-        with gr.Row():
-            with gr.Column(scale=3):
-                down_url_gui = down_url_conf()
-            with gr.Column(scale=1):
-                down_button_gui = down_button_conf()
+            tts_button.click(
+                fn=infer_tts_audio,
+                inputs=[tts_voice, tts_text, tts_active_play],
+                outputs=[aud, tts_play],
+            )
 
-        with gr.Column():
+            # بخش مدل
+            gr.Markdown("### 🎯 Model Selection")
+            
             with gr.Row():
-                model = model_conf()
-                indx = index_conf()
-
-        down_active_gui.change(
-            show_components_down,
-            [down_active_gui],
-            [down_info, down_url_gui, down_button_gui]
-        )
-
-        down_button_gui.click(
-            get_my_model,
-            [down_url_gui],
-            [model, indx]
-        )
+                with gr.Column(scale=1):
+                    model = model_conf()
+                    gr.Markdown("*Upload your .pth model file*")
+                with gr.Column(scale=1):
+                    indx = index_conf()
+                    gr.Markdown("*Upload .index file (optional)*")
+
+            # بخش دانلود از URL
+            down_active_gui = down_active_conf()
+            down_info = gr.Markdown(
+                f"Download models from Hugging Face URLs",
+                visible=False
+            )
+            with gr.Row(visible=False) as url_row:
+                with gr.Column(scale=3):
+                    down_url_gui = down_url_conf()
+                with gr.Column(scale=1):
+                    down_button_gui = down_button_conf()
+
+            down_active_gui.change(
+                show_components_down,
+                [down_active_gui],
+                [down_info, down_url_gui, down_button_gui]
+            )
 
-        with gr.Accordion(label="Advanced settings", open=False):
-            algo = pitch_algo_conf()
-            algo_lvl = pitch_lvl_conf()
-            indx_inf = index_inf_conf()
-            res_fc = respiration_filter_conf()
-            envel_r = envelope_ratio_conf()
-            const = consonant_protec_conf()
-            steps_gui = steps_conf()
-            format_out = format_output_gui()
-            with gr.Row():
-                with gr.Column():
-                    with gr.Row():
-                        denoise_gui = denoise_conf()
-                        effects_gui = effects_conf()
-        button_base = button_conf()
-        output_base = output_conf()
-
-        button_base.click(
-            run,
-            inputs=[
-                aud,
-                model,
-                algo,
-                algo_lvl,
-                indx,
-                indx_inf,
-                res_fc,
-                envel_r,
-                const,
-                denoise_gui,
-                effects_gui,
-                format_out,
-                steps_gui,
-            ],
-            outputs=[output_base],
-        )
+            down_button_gui.click(
+                get_my_model,
+                [down_url_gui],
+                [model, indx]
+            )
 
-        gr.Examples(
-            examples=[
-                [
-                    ["./test.ogg"],
-                    "./model.pth",
-                    "rmvpe+",
-                    0,
-                    "./model.index",
-                    0.75,
-                    3,
-                    0.25,
-                    0.50,
-                ],
-                [
-                    ["./example2/test2.ogg"],
-                    "./example2/model_link.txt",
-                    "rmvpe+",
-                    0,
-                    "./example2/index_link.txt",
-                    0.75,
-                    3,
-                    0.25,
-                    0.50,
-                ],
-                [
-                    ["./example3/test3.wav"],
-                    "./example3/zip_link.txt",
-                    "rmvpe+",
-                    0,
-                    None,
-                    0.75,
-                    3,
-                    0.25,
-                    0.50,
+            # تنظیمات پیشرفته
+            with gr.Accordion(label="⚙️ Advanced Settings", open=False):
+                with gr.Row():
+                    algo = pitch_algo_conf()
+                    algo_lvl = pitch_lvl_conf()
+                
+                with gr.Row():
+                    indx_inf = index_inf_conf()
+                    steps_gui = steps_conf()
+                
+                with gr.Row():
+                    res_fc = respiration_filter_conf()
+                    envel_r = envelope_ratio_conf()
+                    const = consonant_protec_conf()
+                
+                with gr.Row():
+                    format_out = format_output_gui()
+                    denoise_gui = denoise_conf()
+                    effects_gui = effects_conf()
+
+            # دکمه تبدیل
+            button_base = button_conf()
+            
+            # نتیجه
+            gr.Markdown("### 🎵 Output")
+            output_base = output_conf()
+
+            button_base.click(
+                run,
+                inputs=[
+                    aud,
+                    model,
+                    algo,
+                    algo_lvl,
+                    indx,
+                    indx_inf,
+                    res_fc,
+                    envel_r,
+                    const,
+                    denoise_gui,
+                    effects_gui,
+                    format_out,
+                    steps_gui,
                 ],
+                outputs=[output_base],
+            )
 
-            ],
-            fn=run,
-            inputs=[
-                aud,
-                model,
-                algo,
-                algo_lvl,
-                indx,
-                indx_inf,
-                res_fc,
-                envel_r,
-                const,
-            ],
-            outputs=[output_base],
-            cache_examples=False,
-        )
         gr.Markdown(RESOURCES)
 
     return app
 
 
 if __name__ == "__main__":
+    # Get voice list safely
     tts_voice_list = asyncio.new_event_loop().run_until_complete(get_voices_list(proxy=None))
+    
+    # Build voice dropdown options with safe .get() access
     voices = sorted([
-        (" - ".join(reversed(v["FriendlyName"].split("-"))).replace("Microsoft ", "").replace("Online (Natural)", f"({v['Gender']})").strip(), f"{v['ShortName']}-{v['Gender']}")
-        for v in tts_voice_list
+        (
+            " - ".join(
+                reversed(
+                    voice.get("FriendlyName", voice.get("Name", "Unknown Voice")).split("-")
+                )
+            ).replace("Microsoft ", "").replace("Online (Natural)", f"({voice.get('Gender', 'Unknown')})").strip(),
+            f"{voice.get('ShortName', 'Unknown')}-{voice.get('Gender', 'Unknown')}"
+        )
+        for voice in tts_voice_list
     ])
 
+    # Initialize GUI
     app = get_gui(theme)
-
     app.queue(default_concurrency_limit=40)
 
+    # Launch app
     app.launch(
         max_threads=40,
         share=IS_COLAB,
@@ -853,4 +817,4 @@ if __name__ == "__main__":
         quiet=False,
         debug=IS_COLAB,
         ssr_mode=False,
-    )
+    )