Rename =0.2.0 to app.py

app.py

@@ -0,0 +1,114 @@
+import os
+import gradio as gr
+from lib.infer import infer_audio
+from pydub import AudioSegment
+import shutil
+
+# Main function to process audio
+def process_audio(MODEL_NAME, SOUND_PATH, F0_CHANGE, F0_METHOD, MIN_PITCH, MAX_PITCH, CREPE_HOP_LENGTH, INDEX_RATE, 
+                  FILTER_RADIUS, RMS_MIX_RATE, PROTECT, SPLIT_INFER, MIN_SILENCE, SILENCE_THRESHOLD, SEEK_STEP, 
+                  KEEP_SILENCE, FORMANT_SHIFT, QUEFRENCY, TIMBRE, F0_AUTOTUNE, OUTPUT_FORMAT, upload_audio=None):
+
+    # If no sound path is given, use the uploaded file
+    if not SOUND_PATH and upload_audio is not None:
+        SOUND_PATH = os.path.join("uploaded_audio", upload_audio.name)
+        with open(SOUND_PATH, "wb") as f:
+            f.write(upload_audio.read())
+    
+    # Check if a model name is provided
+    if not MODEL_NAME:
+        return "Please provide a model name."
+
+    # Run the inference
+    os.system("chmod +x stftpitchshift")
+    inferred_audio = infer_audio(
+        MODEL_NAME,
+        SOUND_PATH,
+        F0_CHANGE,
+        F0_METHOD,
+        MIN_PITCH,
+        MAX_PITCH,
+        CREPE_HOP_LENGTH,
+        INDEX_RATE,
+        FILTER_RADIUS,
+        RMS_MIX_RATE,
+        PROTECT,
+        SPLIT_INFER,
+        MIN_SILENCE,
+        SILENCE_THRESHOLD,
+        SEEK_STEP,
+        KEEP_SILENCE,
+        FORMANT_SHIFT,
+        QUEFRENCY,
+        TIMBRE,
+        F0_AUTOTUNE,
+        OUTPUT_FORMAT
+    )
+    
+    # Output the inferred audio file
+    return inferred_audio
+
+
+# Gradio Blocks Interface
+with gr.Blocks(tite="Hex RVC") as app:
+    gr.Markdown("## Hex RVC")
+    
+    # Model Input
+    with gr.Row():
+        MODEL_NAME = gr.Textbox(label="Model Name", placeholder="Enter model name")
+
+    # Audio Upload/Input
+    with gr.Row():
+        SOUND_PATH = gr.Textbox(label="Audio Path (Optional)", placeholder="Leave blank to upload audio")
+        upload_audio = gr.File(label="Upload Audio", type='filepath', file_types=["audio"])
+
+    # Main Settings
+    with gr.Row():
+        F0_CHANGE = gr.Number(label="Pitch Change (semitones)", value=0)
+        F0_METHOD = gr.Dropdown(choices=["crepe", "harvest", "mangio-crepe", "rmvpe", "rmvpe+", "fcpe", "fcpe_legacy", 
+                                         "hybrid[mangio-crepe+rmvpe]", "hybrid[mangio-crepe+fcpe]", 
+                                         "hybrid[rmvpe+fcpe]", "hybrid[mangio-crepe+rmvpe+fcpe]"], 
+                                label="F0 Method", value="fcpe")
+
+    # Other Settings
+    with gr.Row():
+        MIN_PITCH = gr.Textbox(label="Min Pitch", value="50")
+        MAX_PITCH = gr.Textbox(label="Max Pitch", value="1100")
+        CREPE_HOP_LENGTH = gr.Number(label="Crepe Hop Length", value=120)
+        INDEX_RATE = gr.Slider(label="Index Rate", minimum=0, maximum=1, value=0.75)
+        FILTER_RADIUS = gr.Number(label="Filter Radius", value=3)
+        RMS_MIX_RATE = gr.Slider(label="RMS Mix Rate", minimum=0, maximum=1, value=0.25)
+        PROTECT = gr.Slider(label="Protect", minimum=0, maximum=1, value=0.33)
+
+    # Advanced Settings
+    with gr.Accordion("Advanced Settings", open=False):
+        SPLIT_INFER = gr.Checkbox(label="Enable Split Inference", value=False)
+        MIN_SILENCE = gr.Number(label="Min Silence (ms)", value=500)
+        SILENCE_THRESHOLD = gr.Number(label="Silence Threshold (dBFS)", value=-50)
+        SEEK_STEP = gr.Slider(label="Seek Step (ms)", minimum=1, maximum=10, value=1)
+        KEEP_SILENCE = gr.Number(label="Keep Silence (ms)", value=200)
+        FORMANT_SHIFT = gr.Checkbox(label="Enable Formant Shift", value=False)
+        QUEFRENCY = gr.Number(label="Quefrency", value=0)
+        TIMBRE = gr.Number(label="Timbre", value=1)
+        F0_AUTOTUNE = gr.Checkbox(label="Enable F0 Autotune", value=False)
+
+    # Output Settings
+    OUTPUT_FORMAT = gr.Dropdown(choices=["wav", "flac", "mp3"], label="Output Format", value="wav")
+
+    # Run Button
+    run_button = gr.Button("Run Inference")
+    
+    # Audio Output
+    output_audio = gr.Audio(label="Generated Audio", type='filepath')
+
+    # Run the process_audio function on button click
+    run_button.click(
+        process_audio, 
+        inputs=[MODEL_NAME, SOUND_PATH, F0_CHANGE, F0_METHOD, MIN_PITCH, MAX_PITCH, CREPE_HOP_LENGTH, INDEX_RATE, 
+                FILTER_RADIUS, RMS_MIX_RATE, PROTECT, SPLIT_INFER, MIN_SILENCE, SILENCE_THRESHOLD, SEEK_STEP, 
+                KEEP_SILENCE, FORMANT_SHIFT, QUEFRENCY, TIMBRE, F0_AUTOTUNE, OUTPUT_FORMAT, upload_audio], 
+        outputs=output_audio
+    )
+
+# Launch the Gradio app
+app.launch()