Update app.py

app.py

@@ -264,10 +264,11 @@ pipe_asr = pipeline(
     batch_size=16,
     torch_dtype=torch_dtype,
     device=device,
-    return_timestamps=True
+    return_timestamps=True,
+    language='en'  # Ensuring transcription is done in English
 )
 
-# Define the transcription function for streaming audio
+
 def transcribe_function(stream, new_chunk):
     try:
         sr, y = new_chunk[0], new_chunk[1]
@@ -275,7 +276,10 @@ def transcribe_function(stream, new_chunk):
         print(f"Error chunk structure: {type(new_chunk)}, content: {new_chunk}")
         return stream, "", None
 
-    y = y.astype(np.float32) / np.max(np.abs(y))
+    y = y.astype(np.float32)
+    max_abs_y = np.max(np.abs(y))
+    if max_abs_y > 0:
+        y = y / max_abs_y
 
     if stream is not None:
         stream = np.concatenate([stream, y])
@@ -289,6 +293,7 @@ def transcribe_function(stream, new_chunk):
 
 
 
+
 # Create the Gradio Blocks interface
 with gr.Blocks(theme="rawrsor1/Everforest") as demo:
     chatbot = gr.Chatbot([], elem_id="RADAR", bubble_full_width=False)