Update app_tsukasa_remote.py

app_tsukasa_remote.py

@@ -135,7 +135,7 @@ with gr.Blocks() as audio_inf:
                 beta = gr.Slider(minimum=0, maximum=1, value=0.0, step=0.1, label="Beta", info="a Diffusion sampler parameter, higher means less affected by the reference | 0 = diffusion is disabled", interactive=True)
                 multispeakersteps = gr.Slider(minimum=3, maximum=15, value=5, step=1, label="Diffusion Steps", interactive=True)
                 embscale = gr.Slider(minimum=1, maximum=5, value=1, step=0.1, label="Intensity", info="will impact the expressiveness, if you raise it too much it'll break.", interactive=True)
-                rate_of_speech = gr.Slider(minimum=0.5, maximum=2, value=1, step=0.1, label="Rate of Speech", info="Higher -> Faster", interactive=True)
+                rate_of_speech = gr.Slider(minimum=0.5, maximum=2, value=1.25, step=0.1, label="Rate of Speech", info="Higher -> Faster", interactive=True)
         
         with gr.Column(scale=1):
             btn = gr.Button("Synthesize", variant="primary")
@@ -174,7 +174,7 @@ with gr.Blocks() as longform:
                                            label="Alpha", 
                                            info="a Diffusion parameter handling the timbre, higher means less affected by the reference | 0 = diffusion is disabled", 
                                            interactive=True)
-                beta_longform = gr.Slider(minimum=0, maximum=1, value=0.3, step=0.1, 
+                beta_longform = gr.Slider(minimum=0, maximum=1, value=0., step=0.1, 
                                           label="Beta", 
                                           info="a Diffusion parameter, higher means less affected by the reference | 0 = diffusion is disabled", 
                                           interactive=True)
@@ -186,7 +186,7 @@ with gr.Blocks() as longform:
                                               info="a Diffusion parameter, it will impact the expressiveness, if you raise it too much it'll break.", 
                                               interactive=True)
 
-                rate_of_speech_longform = gr.Slider(minimum=0.5, maximum=2, value=1, step=0.1, 
+                rate_of_speech_longform = gr.Slider(minimum=0.5, maximum=2, value=1.25, step=0.1, 
                                                     label="Rate of Speech", 
                                                     info="Higher = Faster", 
                                                     interactive=True)
@@ -241,14 +241,14 @@ with gr.Blocks() as prompt_inference:
                                          label="Alpha",
                                          info="a Diffusion sampler parameter handling the timbre, higher means less affected by the reference | 0 = diffusion is disabled",
                                          interactive=True)
-                beta_prompt = gr.Slider(minimum=0, maximum=1, value=0.3, step=0.1,
+                beta_prompt = gr.Slider(minimum=0, maximum=1, value=0., step=0.1,
                                         label="Beta",
                                         info="a Diffusion sampler parameter, higher means less affected by the reference | 0 = diffusion is disabled",
                                         interactive=True)
                 diffusion_steps_prompt = gr.Slider(minimum=3, maximum=15, value=10, step=1,
                                                    label="Diffusion Steps",
                                                    interactive=True)
-                rate_of_speech_prompt = gr.Slider(minimum=0.5, maximum=2, value=1, step=0.1,
+                rate_of_speech_prompt = gr.Slider(minimum=0.5, maximum=2, value=1.25, step=0.1,
                                                   label="Rate of Speech",
                                                   info="Higher = Faster",
                                                   interactive=True)