Update app.py

app.py

@@ -2,6 +2,7 @@ import gradio as gr
 from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
 import torch
 import librosa
+import os
 
 # Model details
 models = {
@@ -31,25 +32,53 @@ def transcribe(audio, model_name):
     transcription = processor.batch_decode(predicted_ids)[0]
     return transcription
 
+# Read HTML banner
+if os.path.exists("banner.html"):
+    with open("banner.html", "r", encoding="utf-8") as file:
+        banner = file.read()
+else:
+    banner = "<h1 style='color: red; text-align: center;'>Banner file not found!</h1>"
+
 # Gradio app
 with gr.Blocks() as demo:
+    gr.HTML(banner)
+
     gr.Markdown("""
     <h1 style="color: #4CAF50; text-align: center;">Persian Speech-to-Text Models</h1>
     <p style="text-align: center;">Test the best Persian STT models in one place!</p>
     """)
 
     with gr.Row():
-        audio_input = gr.Audio(type="filepath", label="Upload your audio file")
-        model_dropdown = gr.Dropdown(
-            choices=list(models.keys()),
-            label="Select Model",
-            value="m3hrdadfi/wav2vec2-large-xlsr-persian-v3"
-        )
+        with gr.Column():
+            audio_input = gr.Audio(
+                source="microphone", type="filepath", label="Upload or Record Audio", 
+                elem_id="audio-upload"
+            )
+            model_dropdown = gr.Dropdown(
+                choices=list(models.keys()),
+                label="Select Model",
+                value="m3hrdadfi/wav2vec2-large-xlsr-persian-v3"
+            )
+
+            # Add Test Audio Button
+            def use_test_audio():
+                return "Test-Audio.ogg"
 
-    output_text = gr.Textbox(label="Transcription", lines=5, placeholder="The transcription will appear here...")
+            test_audio_button = gr.Button("Use Test Audio")
+
+        with gr.Column():
+            output_text = gr.Textbox(
+                label="Transcription", lines=5, placeholder="The transcription will appear here..."
+            )
 
     transcribe_button = gr.Button("Transcribe")
 
+    test_audio_button.click(
+        fn=use_test_audio,
+        inputs=[],
+        outputs=[audio_input]
+    )
+
     transcribe_button.click(
         fn=transcribe,
         inputs=[audio_input, model_dropdown],