Upload folder using huggingface_hub

app.py

@@ -14,11 +14,27 @@ def load_tts_model():
 
 tts = load_tts_model()
 
-def clone(text, audio_file, language, speaking_rate, pitch, volume, 
-          emotion, sample_rate, temperature, seed):
-    if seed is not None:
-        torch.manual_seed(seed)
+# Celebrity voices (example list, you may want to expand or modify this)
+celebrity_voices = {
+    "Morgan Freeman": "path/to/morgan_freeman_sample.wav",
+    "Scarlett Johansson": "path/to/scarlett_johansson_sample.wav",
+    "David Attenborough": "path/to/david_attenborough_sample.wav",
+}
+
+def tts_generate(text, voice, language):
+    with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as temp_audio:
+        temp_audio_path = temp_audio.name
+    
+    tts.tts_to_file(
+        text=text,
+        speaker_wav=celebrity_voices[voice],
+        language=language,
+        file_path=temp_audio_path
+    )
     
+    return temp_audio_path
+
+def clone_voice(text, audio_file, language):
     with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as temp_audio:
         temp_audio_path = temp_audio.name
     
@@ -31,28 +47,48 @@ def clone(text, audio_file, language, speaking_rate, pitch, volume,
     
     return temp_audio_path
 
+# Placeholder function for Talking Image tab
+def talking_image_placeholder():
+    return "Talking Image functionality not implemented yet."
+
 # Define Gradio interface
-iface = gr.Interface(
-    fn=clone,
-    inputs=[
-        gr.Textbox(label="Text"),
-        gr.Audio(label="Voice reference audio file", type="filepath"),
-        gr.Dropdown(["en", "es", "fr", "de", "it"], label="Language", value="en"),
-        gr.Slider(0.5, 2.0, value=1.0, label="Speaking Rate"),
-        gr.Slider(-10, 10, value=0, label="Pitch Adjustment"),
-        gr.Slider(0.1, 2.0, value=1.0, label="Volume"),
-        gr.Dropdown(["neutral", "happy", "sad", "angry"], label="Emotion", value="neutral"),
-        gr.Dropdown([22050, 24000, 44100, 48000], label="Sample Rate", value=24000),
-        gr.Slider(0.1, 1.0, value=0.8, label="Temperature"),
-        gr.Number(label="Seed (optional)")
-    ],
-    outputs=gr.Audio(label="Generated Audio"),
-    title="Advanced Voice Clone",
-    description="Customize your voice cloning experience with various parameters."
-)
+with gr.Blocks() as demo:
+    gr.Markdown("# Advanced Voice Synthesis")
+    
+    with gr.Tabs():
+        with gr.TabItem("TTS"):
+            with gr.Row():
+                tts_text = gr.Textbox(label="Text to speak")
+                tts_voice = gr.Dropdown(choices=list(celebrity_voices.keys()), label="Celebrity Voice")
+                tts_language = gr.Dropdown(["en", "es", "fr", "de", "it"], label="Language", value="en")
+            tts_generate_btn = gr.Button("Generate")
+            tts_output = gr.Audio(label="Generated Audio")
+            
+            tts_generate_btn.click(
+                tts_generate,
+                inputs=[tts_text, tts_voice, tts_language],
+                outputs=tts_output
+            )
+        
+        with gr.TabItem("Talking Image"):
+            gr.Markdown("Talking Image functionality coming soon!")
+        
+        with gr.TabItem("Clone Voice"):
+            with gr.Row():
+                clone_text = gr.Textbox(label="Text to speak")
+                clone_audio = gr.Audio(label="Voice reference audio file", type="filepath")
+                clone_language = gr.Dropdown(["en", "es", "fr", "de", "it"], label="Language", value="en")
+            clone_generate_btn = gr.Button("Generate")
+            clone_output = gr.Audio(label="Generated Audio")
+            
+            clone_generate_btn.click(
+                clone_voice,
+                inputs=[clone_text, clone_audio, clone_language],
+                outputs=clone_output
+            )
 
 # Launch the interface
-iface.launch()
+demo.launch()
 
 # Clean up temporary files (this will run after the Gradio server is closed)
 for file in os.listdir():

requirements.txt

@@ -1,3 +1,3 @@
-gradio
+streamlit
 torch
 TTS