share true

app.py

@@ -25,7 +25,7 @@ def transcribe_and_respond(audio_file):
                 {"role": "system", "content": "Respond naturally and informatively."},
                 {"role": "user", "content": "<|audio|>"}
             ]
-        })
+        }, max_new_tokens=256)
 
         return output
 
@@ -34,12 +34,14 @@ def transcribe_and_respond(audio_file):
 
 # Gradio interface
 with gr.Blocks(title="Shuka v1 Transcription") as iface:
-    gr.Markdown("## 🎙️ Shuka v1 - Voice Transcription")
+    gr.Markdown("## Shuka v1 - Voice Transcription")
+    gr.Markdown("Upload or speak, and the model will respond naturally using SarvamAI's voice foundation model.")
+    
     with gr.Row():
-        audio_input = gr.Audio(sources=["microphone", "upload"], type="filepath", label="Audio")
-        text_output = gr.Textbox(label="Response")
+        audio_input = gr.Audio(sources=["microphone", "upload"], type="filepath", label="Audio Input")
+        text_output = gr.Textbox(label="Model Response", placeholder="Response will appear here...")
 
     audio_input.change(fn=transcribe_and_respond, inputs=audio_input, outputs=text_output)
 
 if __name__ == "__main__":
-    iface.launch(share=True)
+    iface.launch()