Update app.py

app.py

@@ -194,10 +194,10 @@ with gr.Blocks(theme=theme, js=js_func) as clone:
         with gr.Column(scale=1):
             inp = gr.Textbox(label="Text", info="What do you want Vokan to say?", interactive=True)
             voice = gr.Audio(label="Voice", interactive=True, type='filepath', max_length=300, waveform_options={'waveform_progress_color': '#FF593E'})
-            steps = gr.Slider(minimum=3, maximum=60, value=20, step=1, label="Diffusion Steps", info="Higher produces better results typically", interactive=True)
-            embscale = gr.Slider(minimum=1, maximum=10, value=2, step=0.1, label="Embedding Scale", info="Defaults to 2 | low scales may produce unexpected results", interactive=True)
-            alpha = gr.Slider(minimum=0, maximum=1, value=0.3, step=0.1, label="Alpha", info="Defaults to 0.3", interactive=True)
-            beta = gr.Slider(minimum=0, maximum=1, value=0.7, step=0.1, label="Beta", info="Defaults to 0.7", interactive=True)
+            steps = gr.Slider(minimum=3, maximum=100, value=20, step=1, label="Diffusion Steps", info="Higher produces better results typically", interactive=True)
+            embscale = gr.Slider(minimum=1, maximum=10, value=2, step=0.1, label="Embedding Scale", info="Defaults to 2 | High scales may produce unexpected results but may produce more emotional texts", interactive=True)
+            alpha = gr.Slider(minimum=0, maximum=1, value=0.3, step=0.1, label="Alpha", info="Defaults to 0.3 | Resemblance to speakers voice - lower = more similar", interactive=True)
+            beta = gr.Slider(minimum=0, maximum=1, value=0.7, step=0.1, label="Beta", info="Defaults to 0.7 | Resemblance to speakers prosody - lower = more similar - higher = based on sentence", interactive=True)
             speed = gr.Slider(minimum=0.5, maximum=1.5, value=1, step=0.1, label="Speed of speech", info="Defaults to 1", interactive=True)
         with gr.Column(scale=1):
             clbtn = gr.Button("Synthesize", variant="primary")