New approach using pipeline

app.py

@@ -63,6 +63,19 @@ def inference(audio):
 
     return "\n".join(total_prediction) + "\n\n" + ' '.join(words)
 
+pipe = pipeline(
+    task="automatic-speech-recognition",
+    model=model_name,
+    chunk_length_s=10,
+    device=device,
+)
+
+def transcribe(inputs):
+    if inputs is None:
+        raise gr.Error("No audio file submitted! Please upload or record an audio file before submitting your request.")
+
+    text = pipe(inputs, batch_size=1, return_timestamps=True)["text"]
+    return  text
 
 inputs = gr.Audio(label="Input Audio", sources="microphone", type="filepath")
 outputs = "text"
@@ -74,7 +87,7 @@ examples = [
     ["sample2.mp3"],
 ]
 gr.Interface(
-    inference,
+    transcribe,
     inputs,
     outputs,
     title=title,