Update app.py

app.py

@@ -7,6 +7,7 @@ import tempfile
 import os
 import spaces
 
+assert torch.cuda.is_available(), "Demo requires a GPU."
 DEVICE = "cuda" if torch.cuda.is_available() else "cpu"
 print(DEVICE)
 
@@ -52,7 +53,7 @@ with gr.Blocks() as demo:
 
     with gr.Row():
         with gr.Column():
-            gr.Markdown("# Soprano Demo")
+            gr.Markdown("# Soprano Demo\n\nSoprano is an ultra‑lightweight, open‑source text‑to‑speech (TTS) model designed for real‑time, high‑fidelity speech synthesis at unprecedented speed. Soprano can achieve **<15 ms streaming latency** and up to **2000x real-time generation**, all while being easy to deploy at **<1 GB VRAM usage**.")
 
             text_in = gr.Textbox(
                 label="Input Text",
@@ -62,7 +63,7 @@ with gr.Blocks() as demo:
 
             with gr.Accordion("Advanced options", open=False):
                 temperature = gr.Slider(
-                    0.0, 1.0, value=0.3, step=0.05, label="Temperature"
+                    0.0, 1.5, value=0.8, step=0.05, label="Temperature"
                 )
                 top_p = gr.Slider(
                     0.0, 1.0, value=0.95, step=0.01, label="Top-p"
@@ -79,8 +80,8 @@ with gr.Blocks() as demo:
                 autoplay=True,
                 streaming=False,
             )
-            download_btn = gr.Button("Download")
-            file_out = gr.File(label="Download file")
+            #download_btn = gr.Button("Download")
+            #file_out = gr.File(label="Download file")
             gr.Markdown(
                 "Usage tips:\n\n"
                 "- Soprano works best when each sentence is between 2 and 15 seconds long.\n"
@@ -95,11 +96,11 @@ with gr.Blocks() as demo:
         outputs=[audio_out, state_audio],
     )
 
-    download_btn.click(
-        fn=save_audio,
-        inputs=[state_audio],
-        outputs=[file_out],
-    )
+    #download_btn.click(
+    #    fn=save_audio,
+    #    inputs=[state_audio],
+    #    outputs=[file_out],
+    #)
 
 demo.queue()
 demo.launch()